Setting terms of imprisonment

Part 4 Div 1 Crimes (Sentencing Procedure) Act 1999 (ss 44–54, inclusive) contains provisions for setting terms of imprisonment, including non-parole periods, the conditions relating to parole orders, and fixed terms. Different provisions apply depending on whether the court imposes a sentence for a single offence or an aggregate sentence, and whether the offence is in the standard non-parole period Table of Pt 4 Div 1A. Unless the court is imposing an aggregate sentence, it must comply with the requirements of Pt 4 Div 1 by imposing a separate sentence for each offence: s 53(1).

[7-500] Court to set non-parole period

Last reviewed: August 2023

Section 44(1)–(3) Crimes (Sentencing Procedure) Act 1999 provides:


Unless imposing an aggregate sentence of imprisonment, when sentencing an offender to imprisonment for an offence, the court is first required to set a non-parole period for the sentence (that is, the minimum period for which the offender must be kept in detention in relation to the offence).


The balance of the term of the sentence must not exceed one-third of the non-parole period for the sentence, unless the court decides that there are special circumstances for it being more (in which case the court must make a record of its reasons for that decision).


Without affecting the requirement to set a non-parole period for a sentence, a court imposing an aggregate sentence of imprisonment in respect of 2 or more offences on an offender may set one non-parole period for all the offences to which the sentence relates after setting the term of the sentence.


The term of the sentence that will remain to be served after the non-parole period set for the aggregate sentence of imprisonment is served must not exceed one-third of the non-parole period, unless the court decides that there are special circumstances for it being more (in which case the court must make a record of its reasons for that decision).


The court need not indicate the non-parole period that would have been imposed for each offence had separate sentences been imposed instead of an aggregate sentence unless it is required to do so by section 54B.


The failure of a court to comply with subsection (2), (2B) or (2C) does not invalidate the sentence.

Use of “first required to set” in s 44(1) does not mean “determine”

The fact s 44(1) provides that “the court is first required to set a non-parole period” does not mean the non-parole period must first be determined: Musgrove v R [2007] NSWCCA 21 at [44], or that a non-parole period should be set first which is thereafter immutable: R v Way (2004) 60 NSWLR 168 at [111]–[113], citing R v Moffitt (1990) 20 NSWLR 114 with approval; Perry v R [2006] NSWCCA 351 at [14]. It is well established that s 44(1) does not require that the reasoning process begin with the selection of the non-parole period; it is the pronunciation of orders that is required to be done in that way: Eid v R [2008] NSWCCA 255 at [31]. Simpson J added in Musgrove v R at [44] that a literal reading of s 44(1) may lead the court into error:

To determine, initially, the non-parole period, before determining the total sentence, would, in my opinion, (where special circumstances are then found) be conducive to error of the kind exposed in Huynh [[2005] NSWCCA 220]. A finding of special circumstances, after the determination of the non-parole period, would provoke an extension, beyond proper limits, of the balance of term. Sentencing judges need to be wary of taking a course that might lead to that error.

Section 44(1) error in pronouncement of individual sentence

The failure to follow the terms of s 44(1) by pronouncing the non-parole period first and then the balance of term is a technical error which must be corrected: R v Cramp [2004] NSWCCA 264; Itaoui v R [2005] NSWCCA 415 at [17]–[18]; Eid v R [2008] NSWCCA 255 at [31]. If that is the only error, the appellate court should not proceed on the assumption that the exercise of the sentencing discretion miscarried: R v Cramp at [44]; R v Smith [2005] NSWCCA 19 at [10].

Considerations relevant to setting the non-parole period

The non-parole period is imposed because justice requires that the offender serve that period in custody: Muldrock v The Queen (2011) 244 CLR 120 at [57]. It is the minimum period of actual incarceration that the offender must spend in full-time custody having regard to all the elements of punishment including rehabilitation, the objective seriousness of the crime and the offender’s subjective circumstances: Power v The Queen (1974) 131 CLR 623 at 628–629, applied in Deakin v The Queen [1984] HCA 31; R v Simpson (2001) 53 NSWLR 704 at [59]; R v Ogochukwu [2004] NSWCCA 473 at [33]; R v Cramp [2004] NSWCCA 264 at [34]; Caristo v R [2011] NSWCCA 7 at [27]; R v MA [2004] NSWCCA 92 at [34]; Hili v The Queen (2010) 242 CLR 520 at [40]. This principle sets a lower limit to any reduction that might be thought appropriate on the basis of converting punishment into an opportunity for rehabilitation: R v MA at [33].

The risk of re-offending is a relevant factor in setting the minimum term: Bugmy v The Queen (1990) 169 CLR 525 at 537. However, while great weight may be attached to the protection of society in an appropriate case, the sentence imposed should not be more severe than that which would otherwise be appropriate: Veen v The Queen (No 2) (1988) 164 CLR 465 at 477.

The factors relevant to fixing the term of the sentence are the same as the non-parole period, but the weight given to each factor may differ: R v MA at [33]. For example, a serious offence warrants a greater non-parole period due to its deterrent effect upon others, but the nature of the offence does not assume the importance it has when the head sentence is determined: R v MA at [33], citing Bugmy v The Queen at 531–532. Chief Justice Spigelman said of the factor general deterrence in R v Simpson at [64]:

Considerations of general deterrence are at least equally significant to both decisions [fixing the term of the sentence and the non-parole period] which are, in any event, interrelated. Indeed the purport of the High Court’s decision in Power was to reject the proposition that considerations of punishment and deterrence were of primary relevance to the determination of the head sentence and of lesser relevance to the specification of the non-parole period.

In R v Hall [2017] NSWCCA 313, the offender was sentenced to an aggregate sentence of 5 years with a non-parole period of 1 year for historical offences of violence and sexual assault. The judge said the head sentence recognised the objective seriousness of the offences and the non-parole period reflected “considerations of leniency”. That approach was found by the Court of Criminal Appeal to be contrary to the principles in Power v The Queen and R v Simpson: R v Hall at [90].

[7-505] Aggregate sentences

Last reviewed: August 2023

Section 53A(1) Crimes (Sentencing Procedure) Act 1999 enables a court sentencing an offender for multiple offences to impose an aggregate sentence of imprisonment instead of separate individual sentences.

The aggregate sentencing provisions were not intended to create a substantive change to sentencing law: PG v R [2017] NSWCCA 179 at [90]. The scheme was introduced to remove some of the complexity involved when sentencing for multiple offences, while preserving the transparency of the sentencing process. It was intended to overcome the difficulties of applying Pearce v The Queen (1998) 194 CLR 610 and the requirement to set commencement and expiry dates for each sentence: JM v R [2014] NSWCCA 297 at [39]; R v Rae [2013] NSWCCA 9 at [45]; Truong v R [2013] NSWCCA 36 at [231]. The overriding principle is that an aggregate sentence must reflect the totality of the offending behaviour: Burgess v R [2019] NSWCCA 13 at [40]; Aryal v R [2021] NSWCCA 2 at [46]. See [8-220] Totality and sentences of imprisonment.

Section 53A(2) requires a court imposing an aggregate sentence to indicate to the offender, and make a written record of:

  • the fact an aggregate sentence is being imposed: s 53A(2)(a)

  • the sentence that would have been imposed for each offence (after taking into account relevant matters in Pt 3 or any other provision of the Act) had separate sentences been imposed: s 53A(2)(b).

Failure to comply with s 53A does not invalidate an aggregate sentence: s 53A(5).

An aggregate sentence imposed by the Local Court must not exceed 5 years: s 53B.

A court may impose one non-parole period “after setting the term of the [aggregate] sentence” [emphasis added]: s 44(2A).

Use of the word “after” in s 44(2A) is an indication that it is only possible to determine an aggregate non-parole period after deciding the sentence that would have been imposed for each offence. However, failure to comply with s 44(2A) by pronouncing the non-parole period before the total aggregate sentence is a technical error that does not invalidate the sentence: Hunt v R [2017] NSWCCA 305 at [79].

Section 49(2) sets limits as to the duration of the term of an aggregate sentence of imprisonment stating that it:


must not be more than the sum of the maximum periods of imprisonment that could have been imposed if separate sentences of imprisonment had been imposed in respect of each offence to which the sentence relates, and


must not be less than the shortest term of imprisonment (if any) that must be imposed for any separate offence or, if the sentence relates to more than one such offence, must not be less than the shortest term of imprisonment that must be imposed for any of the offences.

The expression in s 49(2)(a) “maximum periods of imprisonment that could have been imposed” appears to mean the maximum penalties for the offences in question. This is based on the text of s 49(1) which provides a single sentence cannot exceed the maximum penalty for the offence.

The aggregate sentence cannot exceed the total of the indicative sentences which should, unless otherwise indicated, be regarded as head sentences for each offence: Dimian v R [2016] NSWCCA 223 at [49]. Indicative sentences should be regarded as head sentences for each of the offences: Dimian v R at [49]. The only circumstance where an indicative sentence might be thought to equate with a non-parole period would be where the sentencing judge expressly states the indicative sentence was to be treated as a fixed term: Dimian v R at [47] with reference to McIntosh v R [2015] NSWCCA 184. See Indicative sentences: fixed term or term of sentence at [7-520].

[7-507] Settled propositions concerning s 53A

Last reviewed: August 2023

In JM v R [2014] NSWCCA 297, RA Hulme J (Hoeben CJ at CL and Adamson J agreeing) at [39], summarised the approach a court should take where it chooses to utilise s 53A:

[39] A number of propositions emerge from the above legislative provisions [ss 44(2C), 53A, 54A(2) and 54B] and the cases that have considered aggregate sentencing:


Section 53A was introduced in order to ameliorate the difficulties of applying the decision in Pearce v The Queen [1998] HCA 57; 194 CLR 610 in sentencing for multiple offences: R v Nykolyn [2012] NSWCCA 219 at [31]. It offers the benefit when sentencing for multiple offences of obviating the need to engage in the laborious and sometimes complicated task of creating a “cascading or ‘stairway’ sentencing structure” when the principle of totality requires some accumulation of sentences: R v Rae [2013] NSWCCA 9 at [43]; Truong v R; R v Le; Nguyen v R; R v Nguyen [2013] NSWCCA 36 at [231]; Behman v R [2014] NSWCCA 239; R v MJB [2014] NSWCCA 195 at [55]–[57].


When imposing an aggregate sentence a court is required to indicate to the offender and make a written record of the fact that an aggregate sentence is being imposed and also indicate the sentences that would have been imposed if separate sentences had been imposed instead (the indicative sentences): s 53A(2). The indicative sentences themselves should not be expressed as a separate sentencing order: R v Clarke [2013] NSWCCA 260 at [50]–[52]. See also Cullen v R [2014] NSWCCA 162 at [25]–[40].


The indicative sentences must be assessed by taking into account such matters in Part 3 or elsewhere in the Crimes (Sentencing Procedure) Act as are relevant: s 53A(2)(b).

There is no need to list such matters exhaustively, but commonly encountered ones in Part 3 include aggravating, mitigating and other factors (s 21A); reductions for guilty pleas, facilitation of the administration of justice and assistance to law enforcement authorities (ss 22, 22A and 23); and offences on a Form 1 taken into account (Pt 3 Div 3). Commonly encountered matters elsewhere in the Act are the purposes of sentencing in s 3A, and the requirements of s 5 as to not imposing a sentence of imprisonment unless a court is satisfied that there is no alternative and giving a further explanation for the imposition of any sentence of 6 months or less.

SHR v R [2014] NSWCCA 94 is an example of a case where a sentencing judge took pleas of guilty into account only in relation to the aggregate sentence, and not in relation to the indicative sentence. This was held (at [42]) to be in breach of the requirement in s 53A(2)(b) …


It is still necessary in assessing the indicative sentences to have regard to the requirements of Pearce v The Queen [1998] HCA 57; 194 CLR 610. The criminality involved in each offence needs to be assessed individually. To adopt an approach of making a “blanket assessment” by simply indicating the same sentence for a number of offences is erroneous: R v Brown [2012] NSWCCA 199 at [17], [26]; Nykolyn v R, supra, at [32]; [56]–[57]; Subramaniam v R [2013] NSWCCA 159 at [27]–[29]; SHR v R, supra, at [40]; R v Lolesio [2014] NSWCCA 219 at [88]–[89]. It has been said that s 53A(2) is “clearly directed to ensuring transparency in the process of imposing an aggregate sentence and in that connection, imposing a discipline on sentencing judges”: [Khawaja v R, [2014] NSWCCA 80] at [18].


The imposition of an aggregate sentence is not to be used to minimise the offending conduct, or obscure or obliterate the range of offending conduct or its totality: R v MJB, supra, at [58]–[60].


One reason why it is important to assess individually the indicative sentences is that it assists in the application of the principle of totality. Another is that it allows victims of crime and the public at large to understand the level of seriousness with which a court has regarded an individual offence: Nykolyn v R, supra, at [58]; Subramaniam v R, supra, at [28]. A further advantage is that it assists when questions of parity of sentencing as between co-offenders arise: R v Clarke, supra, at [68], [75].


Non-parole periods need not be specified in relation to indicative sentences except if they relate to an offence for which a standard non-parole period is prescribed: ss 44(2C) and s 54B(4); AB v R [2014] NSWCCA 31 at [9].


Specification of commencement dates for indicative sentences is unnecessary and is contrary to the benefits conferred by the aggregate sentencing provisions: AB v R, supra, at [10]. Doing so defeats the purpose of a court availing itself of the power to impose an aggregate sentence: Behman v R [2014] NSWCCA 239 at [26]. See also Cullen v R, supra, at [25]–[26].


If a non-custodial sentence is appropriate for an offence that is the subject of the multiple offence sentencing task, it should be separately imposed as was done in Grealish v R [2013] NSWCCA 336. In my respectful view, there was error involved in Behman v R [2014] NSWCCA 239 where an offence with an indicative, but unspecified, non-custodial sentence was included in an aggregate sentence imposed by this Court. The provision for imposing an aggregate sentence in s 53A appears within Part 4 of the Crimes (Sentencing Procedure) Act which is headed “Sentencing procedures for imprisonment”, and within Division 1 of that Part which is headed “Setting terms of imprisonment”.

JM v R has been described as the seminal case explaining the aggregate sentencing scheme: Vaughan v R [2020] NSWCCA 3 at [92]; Taitoko v R [2020] NSWCCA 43 at [130]. However, cases since JM v R elaborate on aspects of the propositions summarised.

Purpose of indicative sentences (proposition 2)

Indicative sentences are required for the purpose of understanding the components of the aggregate sentence in general terms but have no practical operation: Vaughan v R at [90]–[91]; Aryal v R [2021] NSWCCA 2 at [46]. Upon indicating the separate sentences that would have been imposed, the court must then apply the principal of totality to determine an appropriate aggregate sentence: ZA v R [2017] NSWCCA 132 at [70], [74]. There is no requirement to precisely specify any (notional) accumulation of the separate sentences: Vaughan v R at [97]. See further Application of Pearce v The Queen and the totality principle below.

Aggregate sentencing and applying discounts (proposition 3)

Where a court imposes an aggregate sentence it need only explicitly state a discount, or discounts, at the stage of setting each indicative sentence: Glare v R [2015] NSWCCA 194 at [12]; PG v R [2017] NSWCCA 179 at [71], [76]. Where there are multiple offences and the pleas are entered at different times, it is an error to apply an average discount to each indicative sentence: Bao v R [2016] NSWCCA 16 at [44]. All decisions of the court since JM v R are to the effect that a discount must be applied to the starting point of each sentence: for guilty plea discounts see PG v R at [71], [76]; Berryman v R [2017] NSWCCA 297 at [29]; Elsaj v R [2017] NSWCCA 124 at [56]; Ibbotson (a pseudonym) v R [2020] NSWCCA 92 at [138]; for discounts for assistance see TL v R [2017] NSWCCA 308 at [102]–[103].

Application of Pearce v The Queen and the totality principle (propositions 1, 4 and 6)

The principles of sentencing concerning accumulation and concurrency, explained in Pearce v The Queen (1998) 194 CLR 610, do not apply to an aggregate sentence: Vaughan v R [2020] NSWCCA 3 at [91]; Aryal v R [2021] NSWCCA 2 at [46]. However, it is still necessary to consider, albeit intuitively, the extent to which there should be a degree of accumulation between the indicative sentences to arrive at a sentence that reflects the totality of the offending in the particular case: Vaughan v R at [91]; Tuite v R [2018] NSWCCA 175 at [91]; Burgess v R [2019] NSWCCA 13 at [40]; ZA v R [2017] NSWCCA 132 at [70], [74]; Kliendienst v R [2020] NSWCCA 98 at [79]–[102]; see also [8-200] The principle of totality. There is no actual accumulation of the indicative sentences — each offence makes an additional contribution to the totality of the criminality reflected in the aggregate sentence: Aryal v R at [46].

There is no requirement to disclose the precise degree of accumulation between the indicative sentences since that would undermine the legislative purpose of the aggregate sentencing scheme: Berryman v R at [50]; Vaughan v R at [97]; Noonan v R [2021] NSWCCA 35 at [33]. Of this, RA Hulme J said in Vaughan v R, at [117], that:

… a judge does not need to assess a precise degree of accumulation at all [but] simply determines the aggregate sentence by assessing what is appropriate to reflect the totality of criminality in all of the offending. Quite commonly, there are references to there being “notional accumulation” — but if such a reference is apt at all, sight should not be lost of the fact that it is truly something that is “notional”.

Nor is there a requirement, where there are multiple offences committed against multiple complainants, to identify and state by use of “numbers” the notional cumulation internally for each complainant as well as the notional cumulation as between complainants: Benn v R [2023] NSWCCA 24 at [142].

As a result there may be less transparency than when imposing separate sentences: Kliendiest v R at [84]; ZA v R at [88]. Further, the degree of transparency achieved will vary between cases: PW v R [2019] NSWCCA 298 at [6]–[10]. For example, in PW v R, the indicative sentences provided “limited assistance” in understanding the aggregate sentence because the offences were committed in a single, brief episode of criminal conduct where moral culpability and objective seriousness overlapped.

Specifying non-parole periods (proposition 7)

Proposition 7 concerning the requirement to specify a non-parole period for indicative sentences for standard non-parole period offences no longer applies. Since 2016, s 45(1A) Crimes (Sentencing Procedure) Act 1999 permits a sentencing court to decline to set a non-parole period (ie impose a fixed term) for such offences.

Separately imposing a non-custodial sentence (proposition 9)

Proposition 9 was not applied in RL v R [2015] NSWCCA 106 at [63] where the Court of Criminal Appeal said in re-sentencing (for three of the counts) that an “indicative sentence which did not involve a full-time custodial penalty should be adopted”.

Sentencing for backup and related charges

It is permissible to incorporate sentences for related summary offences transferred to the District or Supreme Court pursuant to s 166 Criminal Procedure Act 1986 into a statutory aggregate sentence under s 53A: R v Price [2016] NSWCCA 50 at [76], [80].

Aggregate sentencing and Commonwealth offences

The aggregate sentencing scheme in s 53A can also be used for Commonwealth offenders being sentenced for more than one Commonwealth offence: DPP (Cth) v Beattie [2017] NSWCCA 301 at [146], [210]. However, an aggregate sentence cannot be imposed for a combination of Commonwealth and State offences: Sheu v R [2018] NSWCCA 86 at [26].

See also [16-035] Sentencing for multiple offences.

For appeals against aggregate sentences see: [70-035] Appellate review of an aggregate sentence and Aggregate sentences in [70-090] Purpose and limitations of Crown appeals.

[7-510] Special circumstances under s 44(2) or (2B)

Last reviewed: August 2023

Section 44(2) and (2B) Crimes (Sentencing Procedure) Act 1999 provide that the non-parole period for either a single sentence or an aggregate sentence must not fall below three-quarters of the term of the sentence unless there is a finding of special circumstances (in which case reasons must be recorded for the decision).

In R v GDR (1994) 35 NSWLR 376 at 381, a five-judge Bench said, after noting the limit of the restriction in the former s 5(2) Sentencing Act 1989 (the statutory predecessor of s 44(2)):

In practice, the principles of general law to which reference has been made, and which affect the relationship between a minimum and an additional term, may well operate to produce the result that, in many cases, the additional term will be one-third of the minimum term, for the reason that the sentencing judge considers that the period available to be spent on parole should be not less than one-quarter of the total sentence. What was said in Griffiths [(1989) 167 CLR 372] about the pattern of sentencing in this State before the enactment of the legislation there referred to suggests that this will frequently be so. That does not mean, however, that sentencing judges have been deprived, by s 5, of their discretion. It is, rather, the consequence of the fact that in many cases a proper exercise of discretion will dictate that the additional term be not less than one-third of the minimum term, or one-quarter of the total sentence. In a practical sense, therefore, in many cases, the result will be an additional term which is one-third of the minimum term. This will be because the statute says it cannot be more (in the absence of special circumstances), and because general sentencing principles dictate, in the particular case, that it should not be less [emphasis added].

The language of s 44(2) constrains or fetters the sentencing discretion by providing that the balance of term must not exceed the non-parole period by one-third unless the court finds special circumstances.

Balance of term in excess of one-third

There is no corresponding rule that the balance of term must not be less than one-third of the non-parole period: Musgrove v R [2007] NSWCCA 21 at [27]; DPP (NSW) v RHB [2008] NSWCCA 236 at [17], [19]; Wakefield v R [2010] NSWCCA 12 at [26]. However, it is advisable for the court to explain why a ratio in excess of 75% was selected to avoid an inference that an oversight must have occurred: Wakefield v R at [26]; Briggs v R [2010] NSWCCA 250 at [34] cited in Russell v R [2010] NSWCCA 248 at [41]; Etchell v R [2010] NSWCCA 262 at [49]–[50]; Maglovski v R [2014] NSWCCA 238 at [28]; Brennan v R [2018] NSWCCA 22 at [69]. An express comment is preferable because it makes clear the judge is aware of the impact of any accumulation: GP v R [2017] NSWCCA 200 at [22]. This is more than simply a salutary discipline; offenders should not be left to wonder whether the term of their incarceration was affected by inadvertent oversight or whether it was fully intended: Huang v R [2019] NSWCCA 144 at [52]. For example, the judge’s silence in Briggs v R left “a sense of disquiet that he may have overlooked giving appropriate focus to the statutory ratio”: per Fullerton J at [34]; see also Huang v R at [53] and Hardey v R [2019] NSWCCA 310 at [34]. This is especially the case where consecutive sentences are imposed: Dunn v R [2007] NSWCCA 312. The reasons do not need to be lengthy. In Brennan v R, the judge gave “short but adequate reasons” for imposing a non-parole period greater than 75%: at [40].

Even in circumstances where there is no specific reference to the requirements of s 44(2), consideration of the reasons as a whole may indicate there was no oversight. For example, in Sonter v R [2018] NSWCCA 228 at [23], the court found that although there was no specific reference to the ratio between the non-parole period and the head sentence, a number of factors identified by the judge during his reasons, including a specific reference to the need to have regard to totality, overwhelmingly pointed to a conclusion that no oversight had occurred.

Nonetheless, imposing a non-parole period greater than 75% is an adverse and exceptional outcome in NSW sentencing practice: Brennan v R at [72]–[90]. As a matter of procedural fairness, where a judge is considering whether to impose a non-parole period greater than 75%, the particular circumstances of the case may require the judge to invite submissions from the parties on the topic: Brennan v R at [96]–[97].

Section 44(2) and (2B) only require reasons to be given if a finding of special circumstances is made: Rizk v R [2020] NSWCCA 291 at [138]–[139]. However, it is also advisable to do so where such a finding is not made to avoid an inference the matter was not considered: Maglovski v R at [28]; Calhoun (a pseudonym) v R [2018] NSWCCA 150 at [30].

[7-512] Special circumstances generally

Last reviewed: August 2023

Parliament has not prescribed at which stage of the sentencing exercise the court must consider the issue of special circumstances. There is nothing in s 44 Crimes (Sentencing Procedure) Act 1999 or the case law which mandates a method or, to adopt the High Court’s term in Markarian v The Queen (2005) 228 CLR 357 at [27], the “path” the court must take.

See What constitutes special circumstances? (at [7-514] below) as to the factors that may be relevant in a particular case. An offender’s legal representative is expected to make submissions addressing factors which may warrant a finding of special circumstances and particularly what is an appropriate period of supervision on parole for the offender: Edwards v R [2009] NSWCCA 199 at [11]; Jinnette v R [2012] NSWCCA 217 at [96].

If there are circumstances that are capable of constituting special circumstances, the court is not obliged to vary the statutory ratio. Before a variation is made “it is necessary that the circumstances be sufficiently special”: R v Fidow [2004] NSWCCA 172 at [22]; Langbein v R [2013] NSWCCA 88 at [54]. The decision is — first, one of fact, to identify the circumstances, and secondly, one of judgment — to decide whether the circumstances justify a lowering of the non-parole period below the statutory ratio: R v Simpson (2001) 53 NSWLR 704 at [73]; Fitzpatrick v R [2010] NSWCCA 26 at [36].

A finding of special circumstances is a discretionary finding of fact: R v El-Hayek [2004] NSWCCA 25 at [103]; Caristo v R [2011] NSWCCA 7 at [28].

A finding of special circumstances permits an adjustment downwards of the non-parole period, but it does not authorise an increase in the term of the sentence: R v Tobar [2004] NSWCCA 391 at [36]–[37]; R v Huynh [2005] NSWCCA 220 at [35]–[39]; Markham v R [2007] NSWCCA 295 at [29]. As with the statutory predecessor (s 5(2) Sentencing Act 1989 (rep)), ss 44(2) and 44(2A) should not be understood as statutory norms (75% or 3:1) in the sense that variation in either direction, up or down, absent special circumstances is contrary to law: R v GDR (1994) 35 NSWLR 376 at 380. The extent of the adjustment is not determined by any “norm” and the court is to be guided by general sentencing principles: Caristo v R at [28].

In setting an effective non-parole period for more than one offence the focus should not be solely upon the percentage proportions that the non-parole periods have to the total term. In Caristo v R, RA Hulme J said at [42]: “The actual periods involved are equally, and probably more, important.”

When a court decides to reduce the non-parole period because of a finding of special circumstances, double counting matters already taken into account in calculating the head sentence should be avoided: R v Fidow at [18]; Trindall v R [2013] NSWCCA 229 at [17]; Langbein v R at [54]; Ho v R [2013] NSWCCA 174 at [33].

The degree or “extent of any adjustment to the statutory requirement is essentially a matter within the sentencing judge’s discretion”: Clarke v R [2009] NSWCCA 49 at [13]; R v Cramp [2004] NSWCCA 264 at [31]) including consideration of those circumstances which concern the nature and purpose of parole: R v GDR at 381.

Although the desirability of an offender undergoing suitable rehabilitative treatment is capable of being a special circumstance, where special circumstances are found on this basis, it is an error for a court to refrain from adjusting the sentence based on a view that the offender would benefit from treatment while in full-time custody: Muldrock v The Queen (2011) 244 CLR 120 at [57]–[58]. This is because full-time custody is punitive and treatment in prison is a matter in the executive’s discretion. Also, an offender may not qualify for a program in custody or it may not be available: Muldrock v The Queen at [57].

A court can have regard to the practical limit of 3 years on parole supervision which an offender may receive under cl 214A Crimes (Administration of Sentences) Regulation 2014. With regard to the operation of cl 228 Crimes (Administration of Sentences) Regulation 2008 (rep), which was in similar terms to cl 214A, see the discussion in: AM v R [2012] NSWCCA 203 at [90]; Collier v R [2012] NSWCCA 213 at [37]; Jinnette v R at [107]. However, cl 214A provides in the case of a “serious offender” (defined in s 3(1) Crimes (Administration of Sentences) Act 1999) that the period of supervision may be extended by, or a further period of supervision imposed of, up to 3 years at a time.

A purported failure to adjust a sentence for special circumstances raises so many matters of a discretionary character that the Court of Criminal Appeal has been reluctant to intervene. The court will only intervene if the non-parole period is manifestly inadequate or manifestly excessive: R v Cramp [2004] NSWCCA 264 at [31]; R v Fidow at [19]; Jiang v R [2010] NSWCCA 277 at [83]. Ultimately the non-parole period that is set is what the court concludes, in all of the circumstances, ought to be the minimum period of incarceration: Muldrock v The Queen at [57]; R v Simpson at [59].

[7-514] What constitutes special circumstances?

Last reviewed: August 2023

The full range of subjective considerations is capable of warranting a finding of special circumstances: R v Simpson (2001) 53 NSWLR 704 at [46], [60]. It will be comparatively rare for an issue to be incapable, as a matter of law, of ever constituting a “special circumstance”: R v Simpson at [60]. Findings of special circumstances have become so common that it appears likely that there can be nothing “special” about many cases in which the finding is made: R v Fidow [2004] NSWCCA 172 at [20].


Generally speaking, the reform of the offender will often be the purpose in finding special circumstances, but this is not the sole purpose: R v El-Hayek [2004] NSWCCA 25 at [105]. In Kalache v R [2011] NSWCCA 210 at [2], Allsop P recognised that the concept of special circumstances “bears upon an important element and purpose of the sentencing process, rehabilitation”. However, the incongruity of tying s 44(2) Crimes (Sentencing Procedure) Act to rehabilitation was observed by Spigelman CJ in R v Simpson (2001) 53 NSWLR 704 at [58]:

… the requirements of rehabilitation would be best computed in terms of a period of linear time, not in terms of a fixed percentage of a head sentence. The desirability of a longer than computed period of supervision will be an appropriate approach in many cases.

Nevertheless, an offender’s good prospects of rehabilitation may warrant a finding of special circumstances: Arnold v R [2011] NSWCCA 150 at [37]; RLS v R [2012] NSWCCA 236 at [120]. It is not necessary to be satisfied rehabilitation is likely to be successful as opposed to a possibility, but merely that the offender has prospects of rehabilitation which would be assisted by a longer parole period: Thach v R [2018] NSWCCA 252 at [45]–[46]. However, if an offender has poor prospects of rehabilitation and shows a lack of remorse, protection of the society may assume prominence in the sentencing exercise and militate against a finding of special circumstances: R v Windle [2012] NSWCCA 222 at [55].

Risk of institutionalisation

The risk of institutionalisation, even in the face of entrenched and serious recidivism, may justify a finding of special circumstances: Jackson v R [2010] NSWCCA 162 at [24]; Jinnette v R [2012] NSWCCA 217 at [103]. However, the existence of the factor does not require a finding: Dyer v R [2011] NSWCCA 185 at [50]; Jinnette v R at [98]. If institutionalisation has already occurred, the focus may be on ensuring that there is a sufficient period of conditional and supervised liberty to ensure protection of the community and to minimise the chance of recidivism: Jinnette v R at [103].

Drug and alcohol addiction

A finding of special circumstances may be made where the offender requires substantial help to overcome drug and alcohol addiction: Sevastopoulos v R [2011] NSWCCA 201 at [84]–[85]; or where there is a recognition of an offender’s efforts to rehabilitate himself or herself from drug addiction and a demonstrated need for continued assistance if those efforts are to be maintained: R v Vera [2008] NSWCCA 33 at [20].

First custodial sentence

It is doubtful whether the fact a sentence represents an offender’s first time in custody may alone justify finding special circumstances: Collier v R [2012] NSWCCA 213 at [36]; Singh v R [2020] NSWCCA 353 at [79]; R v Kaliti [2001] NSWCCA 268 at [12]; R v Christoff [2003] NSWCCA 52 at [67]; Langbein v R [2008] NSWCCA 38 at [112]; Clarke v R [2009] NSWCCA 49 at [12]. Although such a finding may be made in combination with other factors: Leslie v R [2009] NSWCCA 203 at [37]; R v Little [2013] NSWCCA 288 at [30].

Ill health, disability or mental illness

There are many examples in which ill health, mental illness or a disability are found to be circumstances which may contribute to a finding of special circumstances: R v Sellen (unrep, 5/12/91, NSWCCA); R v Elzakhem [2008] NSWCCA 31 at [68]; Muldrock v The Queen (2011) 244 CLR 120 at [58]; Devaney v R [2012] NSWCCA 285 at [92]; Morton v R [2014] NSWCCA 8 at [19].

Accumulation of individual sentences

There is a conventional sentencing practice of finding special circumstances in cases where sentences imposed for multiple offences are served consecutively in order to apply the totality principle: Hejazi v R [2009] NSWCCA 282 at [36]. Sentencing judges are required to give effect to the principle of totality and therefore should have regard to the outcome of any such accumulation: R v Simpson (unrep, 18/6/92, NSWCCA); R v Close (1992) 31 NSWLR 743 at 748–749; R v Clarke (unrep, 29/3/95, NSWCCA); R v Clissold [2002] NSWCCA 356 at 19], [21]; Cicekdag v R [2007] NSWCCA 218 at [49]; R v Elzakhem [2008] NSWCCA 31 at [68]–[69]; Hejazi v R at [35]. However, in Singh v R at [77]–[79], RA Hulme J (Johnson J agreeing) observed that the rationale for finding special circumstances identified in Simpson v R did not apply when an aggregate sentence was imposed.

An accumulation of sentences does not automatically give rise to a finding that special circumstances exist: R v Cook [1999] NSWCCA 234 at [38]. Where the court utilises the power to impose an aggregate sentence under s 53A, the issue of special circumstances is governed by s 44(2B): see Aggregate sentences at [7-505].

Protective custody

A court cannot find special circumstances on account of protective custody unless the offender provides evidence that his or her conditions of incarceration will be more onerous than usual: RWB v R [2010] NSWCCA 147 at [192]–[195]; Langbein v R [2008] NSWCCA 38 at [113] and cases cited therein: Mattar v R [2012] NSWCCA 98 at [23]–[25].

Care should be taken to avoid counting hardship of protective custody as a reason for discounting the total sentence and again as a factor establishing special circumstances: R v S [2000] NSWCCA 13 at [33]; R v Lee [2000] NSWCCA 392 at [80].

Similarly, where an offender has been given a generous discount on the head sentence for providing assistance to authorities (partly because of the resulting need to serve the sentence in protection) it is not then permissible to make a finding of special circumstances on the basis that the sentence will be served in virtual solitary confinement: R v Capar [2002] NSWCCA 517 at [28]–[29].

See Hardship of custody at [10-500] and Hardship of custody for child sex offender at [17-570] Mitigating factors.


An offender’s youth is a common ground for a finding of special circumstances: Hudson v R [2007] NSWCCA 302 at [6]; MB v R [2007] NSWCCA 245 at [23]; R v Merrin [2007] NSWCCA 255 at [55]; Kennedy v R [2008] NSWCCA 21 at [53]; AM v R [2012] NSWCCA 203 at [86].

Advanced age may similarly be a factor: R v Mammone [2006] NSWCCA 138 at [54].

Hardship to family members

Hardship to members of an offender’s family is generally irrelevant and can only be taken into account in highly exceptional circumstances: King v R [2010] NSWCCA 202 at [18], [23], [25]. The care of young children is not normally an exceptional circumstance: R v Murphy [2005] NSWCCA 182 at [16]–[19].

However, in R v Grbin [2004] NSWCCA 220 at [33], special circumstances were found where there was evidence of the importance of the strong bond between the offender and his son, who suffered from clinical autism and other disabilities and required constant supervision. See also R v Maslen (unrep, 7/4/95, NSWCCA) where the child was severely disabled and R v Hare [2007] NSWCCA 303 where the child suffered from Asperger’s Syndrome.

A finding that the offender has good prospects for rehabilitation and is a mother of a young child, may support a finding of special circumstances: R v Bednarz [2000] NSWCCA 533 at [13], [52] (a two-judge bench case referred to in Harrison v R [2006] NSWCCA 185 at [31]); R v Gip [2006] NSWCCA 115 at [28]–[30], [68].


Special circumstances may be found where there is a degree of self-inflicted shame and guilt already suffered combined with a mental condition: R v Dhanhoa [2000] NSWCCA 257 at [16], [45]; R v Koosmen [2004] NSWCCA 359 at [34]; R v Elkassir [2013] NSWCCA 181 at [37]. However, the weight attributed to the factor cannot lead to the imposition of an inadequate non-parole period: R v Elkassir at [73]. Where the facts reveal gross moral culpability, judges should be wary of attaching too much weight to considerations of self-punishment. Genuine remorse and self-punishment do not compensate for, or balance out, gross moral culpability: R v Koosmen at [32].


The need in a particular case to preserve proper parity between co-offenders may itself amount to special circumstances but such an application of s 44(2) must be justified by the special requirements of a particular sentencing exercise: Tatana v R [2006] NSWCCA 398 at [33]; Briouzguine v R [2014] NSWCCA 264 at [67]. Generally disparity will not arise simply because the application of s 44 to particular offenders results in different sentences between co-offenders: R v Do [2005] NSWCCA 209 at [18]–[19]; Gill v R [2010] NSWCCA 236 at [60]–[62].

Sentencing according to past practices

Sentencing according to past practices may justify a finding of special circumstances in order to reflect the applicable non-parole period/head sentence ratio at the time: AJB v R [2007] NSWCCA 51 at [36]–[37]; MJL v R [2007] NSWCCA 261 at [42].

See Sentencing for historical child sexual offences at [17-410].

[7-516] Giving effect to finding of special circumstances

Last reviewed: November 2023

Where a finding of special circumstances is expressed for an individual sentence or individual sentences, the ultimate sentence imposed should usually give effect to that finding unless there are express reasons for not doing so.

The Crimes (Sentencing Procedure) Act 1999 contains no express requirement for a judge to apply the statutory ratio to an effective or overall sentence, but s 44(2) has been found to apply in that situation and also where a sentence is accumulated on an existing sentence: Lonsdale v R [2020] NSWCCA 267 at [65]; GP v R [2017] NSWCCA 200 at [16]; Harris v R [2023] NSWCCA 44 at [19], [30]; Rizk v R [2020] NSWCCA 291 (which also considers s 44(2B)).

While s 44(2) does not directly require a judge to give reasons for setting a non-parole period exceeding 75% of the total or effective sentence, it is advisable to do so: Lonsdale v R at [31]; [65]; GP v R at [22]; CM v R [2013] NSWCCA 341 at [39]. However, this does not require the performance of a mathematical calculation to the determination of the proportion of the non-parole period to a total term where a particular sentence is accumulated on an existing sentence: Lonsdale v R at [32]; Zreika v R [2020] NSWCCA 345 at [26].

On appeal, determining whether the lack of adjustment of the statutory ratio reflected in the overall term is intentional or the result of inadvertence or miscalculation often depends on what can be gleaned of the judge’s intention from the sentencing remarks: CM v R at [40]; Maglis v R at [24]; Harris v R [2023] NSWCCA 44 at [19]. In CM v R there was nothing to indicate that the judge was aware of, or intended, the final result and so the ground that the judge failed to give practical effect to the finding of special circumstances in the total effective sentence was upheld: CM v R at [42]. In AB v R [2014] NSWCCA 31, even though the judge’s finding of special circumstances was not reflected in the overall sentence, the final result was what the judge intended and there was no inadvertence or miscalculation: at [54], [57]; see also Sampson v R [2023] NSWCCA 239 at [6]–[13]. Similarly, in Rizk v R at [143], [146] and Lonsdale v R at [39], the particular sentencing judges did not err by not giving express reasons for imposing an effective non-parole period that exceeded 75%, to a modest degree.

On the other hand, the court found error in Sabongi v R [2015] NSWCCA 25, where the sentencing judge failed to give effect to an intention to vary the overall ratio to take account of the applicant’s mental condition, the need for rehabilitation and supervision, and the accumulation of sentences. See also Woods v R [2020] NSWCCA 219 at [71], [73].

The focus of the inquiry should not be solely upon the percentage proportions that the non-parole and parole periods bear to the total term. The actual periods involved are equally, and probably more, important: Woods v R at [62]; MD v R [2015] NSWCCA 37 at [41]; Caristo v R [2011] NSWCCA 7 at [42]. Care may be required when an applicant is sentenced in NSW while serving a sentence in another State where the statutory ratio of non-parole period to sentence may vary: see, for example, Ozan v R [2021] NSWCCA 231.

The Sentencing calculator on JIRS may assist when considering the requirements of s 44.

[7-518] Empirical study of special circumstances

Last reviewed: August 2023

A 2013 study by the Judicial Commission examined sentencing cases finalised in the NSW District and Supreme Courts for the period 1 January 2005 to 30 June 2012: P Poletti and H Donnelly, “Special circumstances under s 44 of the Crimes (Sentencing Procedure) Act 1999”, Sentencing Trends & Issues, No 42, Judicial Commission of NSW, 2013.

An analysis of the sentencing statutes of other Australian jurisdictions revealed that NSW is one of few jurisdictions with a statutory rule which constrains a court’s discretion when it sets a non-parole period. Further, the ratio set in s 44(2) and s 44(2A) Crimes (Sentencing Procedure) Act 1999 is comparatively high.

Special circumstances were found in the vast majority of cases (91.4%) and was found more frequently for the youngest offenders (98.8% for juveniles and 96.8% for offenders aged 18–20 years) and for the oldest offenders (100% for offenders aged over 70 years and 98.0% for offenders aged 66–70 years).

A random sample of 159 judgments was examined. The most common reasons for finding special circumstances was the offender’s need for a lengthy period of supervision in the community after release (66.7%), followed by the lack of a prior criminal record (35.8%). These common reasons mostly referred to the offender serving their first prison sentence. Other common reasons include good prospects of rehabilitation (29.6%), age of the offender — particularly youth (25.8%), the effect of accumulation (23.3%) and hardship of custody (10.1%). The reasons given should not be viewed in isolation as there is a clear interrelationship between the different reasons.

The study (see table 3 in the study) analysed mean ratios for the basic and aggravated forms of robbery, break and enter, sexual assault and the supply of a prohibited drug. Subject to one (explicable) exception, the authors found that the longer the sentence and the more serious the crime, the lower the frequency of finding special circumstances. This is because for longer sentences the period of supervision was considered sufficient without a finding of special circumstances. More serious offences (such as murder and aggravated sexual assault in company) recorded the lowest frequency of special circumstances, which was unsurprising given the longer duration of their sentences and the limited utility of an extended period of supervision.

[7-520] Court may decline to set non-parole period

Last reviewed: August 2023

Section s 45(1) Crimes (Sentencing Procedure) Act 1999 provides:

When sentencing an offender to imprisonment for an offence, or in the case of an aggregate sentence of imprisonment, for offences, a court may decline to set a non-parole period for the offence or offences if it appears to the court that it is appropriate to do so:


because of the nature of the offence to which the sentence, or of each of the offences to which an aggregate sentence relates, or the antecedent character of the offender, or


because of any other penalty previously imposed on the offender, or


for any other reason that the court considers sufficient.

Section 45(1A) permits a court to decline to set a non-parole period (ie, impose a fixed term) for an offence to which a standard non-parole period applies. Section 45(1A) does not apply to sentencing for an offence dealt with summarily or if the offender is under 18 years of age: s 45(1B).

Where the court declines to set a non-parole period, it must make a record of its reasons for declining to do so: s 45(2). R v Parsons [2002] NSWCCA 296 and Collier v R [2012] NSWCCA 213 at [55] are examples of cases where the sentencing judge erred by not fixing a non-parole period and not giving reasons as to why he declined to do so. The discretion in s 45(1), construed literally, is simply a discretion to decline to set a non-parole period: Collier v R at [58]. However, the weight of authority (both in relation to s 45(1) and its statutory predecessor under s 6 Sentencing Act 1989) supports the view that where a fixed term is imposed it should be set at an equivalent level, or equate to, what the non-parole period would have been: Collier v R at [56]–[58], citing R v Dunn [2004] NSWCCA 346 at [161]. The question whether s 45(1) also permits a court to impose a fixed term to reduce an otherwise appropriate sentence may be a future topic for resolution: Collier v R at [62]; see further below.

When sentencing an offender for multiple offences and where some accumulation is appropriate (assuming the aggregate sentence provision is not utilised), it is acceptable to impose fixed terms of imprisonment for some or most of the sentences. This is because, if a sentence containing a non-parole period and a parole period were set for each offence, the parole terms of many of these sentences would be subsumed in the non-parole period or fixed term of some longer sentence(s): R v Dunn at [161]. The judge in R v Burgess [2005] NSWCCA 52 decided that parole supervision would not be of any benefit to the offenders and imposed a fixed term under s 45(1): at [45].

For further discussion see Concurrent and consecutive sentences at [8-200]ff.

Indicative sentences: fixed term or term of sentence?

There is controversy as to whether or not an indicative sentence equates to a fixed term and whether a fixed term should be equated with a non-parole period. The divergent authority was summarised by N Adams J in Waterstone v R [2020] NSWCCA 117 at [62]–[73]. Although it did not arise in the appeal, her Honour observed that she doubted whether a fixed term should be equated with a non-parole period: at [81]–[90]; cf Johnson J at [4]ff.

In McIntosh v R [2015] NSWCCA 184, where the appeal concerned an aggregate sentence, the court (Basten JA, Wilson J agreeing; Hidden J dissenting on this point) held that where a sentence is indicated under s 53A(2)(b) for an offence that is not subject to a standard non-parole period, it is permissible to indicate a fixed term (or mandatory period of custody). Basten JA at [166]–[167] followed R v Dunn. His Honour held that there is nothing in the language of ss  44 and 45 which denies the court the power to approach the indication of a sentence under s 53A(2) in the manner described in R v Dunn and, unless there are compelling reasons to the contrary, R v Dunn should be followed: at [167].

Hidden J did not agree. In his Honour’s view, the total term (or head sentence) for each offence should be indicated, not the minimum period of mandatory custody. The head sentence reflects the assessment of criminality of an offence taking into account all the relevant circumstances and it is that assessment which should be reflected in an indicative sentence: at [173], [174].

The approach taken by the court in McIntosh v R in relation to fixed terms and indicative sentences was the subject of comment in (2015) 22(8) CrimLN 127 at [3572] where it was argued that the “fixed term” indicative sentence approach begs error because it, inter alia, “may lead a court into error in not having regard to the full sentence for an offence in comparison to its maximum penalty” and prevents the community, particularly victims, from being informed “of the court’s sentencing response to an individual offence”. It is to be also noted that it is permissible under s 45(1) for a court to impose an aggregate fixed term sentence.

Subsequently in Dimian v R [2016] NSWCCA 223 at [46] the court held that on any proper construction of s 53A(2), seen in the context of the whole Act, the “sentence that would have been imposed” must be a reference to the overall, or term, of sentence. Any suggestion that an indicative sentence is the non-parole period is inconsistent with the principles of aggregate sentencing set out in JM v R [2014] NSWCCA 297 at [39]: Dimian v R at [47]. The only circumstance where an indicative sentence might be thought to equate with a non-parole period would be where the sentencing judge expressly states that the indicative sentence was to be treated as a fixed term: Dimian v R at [47]. In Dimian v R, the court found the judge erred by imposing an aggregate sentence which exceeded the sum of the indicative sentences: at [49].

[7-530] Court not to set non-parole period for sentence of 6 months or less

Last reviewed: August 2023

Section 46 Crimes (Sentencing Procedure) Act 1999 provides that a court may not set a non-parole period for a sentence of imprisonment if the term of the sentence is 6 months or less. Section 46(1) does not apply if a court imposes an aggregate sentence of imprisonment in respect of two or more offences of more than six months, even if the individual sentences the court would have imposed would have been less than six months (as referred to in s 53A(2)(b)): s 46(2).

If the court decides to set a term of imprisonment of 6 months or less, then it must make a record of its reasons for doing so, including its reasons for deciding: that no penalty other than imprisonment is appropriate; and not to allow the offender to participate in an intervention program or other program for treatment and rehabilitation: s 5(2) Crimes (Sentencing Procedure) Act.

[7-540] Commencement of sentence

Last reviewed: August 2023

The law relating to commencement of sentence is set out in s 47 Crimes (Sentencing Procedure) Act 1999. In summary, every sentence or aggregate sentence passed takes effect from the time it is passed, unless the court otherwise directs. Thus, if the sentencer does not specify the date for commencement, it will be deemed to commence on the day on which the sentence or aggregate sentence was imposed. This section confers power to direct that a sentence may commence upon any determinate date either subsequent or prior to the time when it was imposed. Subject to a statutory provision(s) to the contrary, a sentence of imprisonment runs from the date it is imposed: Whan v McConaghy (1984) 153 CLR 631 at 636; R v Hall [2004] NSWCCA 127 at [28]; Kaderavek v R [2018] NSWCCA 92 at [19]. If the sentence commences before the date the sentence is imposed, s 47 provides no guidance except that the sentencing judge “must take into account any time for which the offender has been held in custody in relation to the offence”. If the sentence commences after that date, there is less flexibility as a result of s 47(4) and s 47(5): Kaderavek v R at [19].

On the issues of:

  • how to count pre-sentence custody and the necessity of backdating see [12-500] Counting pre-sentence custody

  • forward dating sentences of imprisonment see [7-547]

  • what time should be counted including offences committed whilst the offender was on parole see [12-510] What time should be counted?

  • taking into account participation of the offender in intervention programs see [12-520] Intervention programs

  • quasi-custody bail conditions such as the MERIT program see [12-530] Quasi-custody bail conditions

  • having regard to the fact the offender will be serving his or her sentence in protective custody see [10-500] Hardship of custody.

[7-545] Rounding sentences to months

Last reviewed: August 2023

The court in Rios v R [2012] NSWCCA 8 raised the issue of rounding and whether a sentence should be expressed in terms of years, months and days, as opposed to just years and months. Adamson J said at [43] with reference to Ruano v R [2011] NSWCCA 149 at [20] that expressing a sentence with days “… ought be discouraged because it adds an unnecessary complication in the sentencing process”. In appropriate cases an adjustment should be made by rounding the number of days down to a number of months: Rios v R at [43].

[7-547] Forward dating sentences of imprisonment

Last reviewed: August 2023

Section 47(2)(b) Crimes (Sentencing Procedure) Act 1999 provides that a court may direct that a sentence of imprisonment commences “on a day occurring after the day on which the sentence is imposed, but only if the sentence is to be served consecutively (or partly concurrently and partly consecutively) with some other sentence of imprisonment”.

Section 47(5) provides that a direction under s 47(2)(b) may not be made in relation to a sentence of imprisonment imposed on an offender who is serving some other sentence of imprisonment by way of full-time detention if:


a non-parole period has been set for that other sentence, and


the non-parole period for that other sentence has expired, and


the offender is still in custody under that other sentence.

Section 47(5) governs a specific scenario where the offender is still in custody under what is described as the “other sentence”. It is a statutory rule as to when the second sentence must commence where the statutory criteria are met. If the criteria in s 47(5) apply, the court does not have the power to impose a sentence in the terms of s 47(2)(b) “on a day occurring after the day on which the sentence is imposed”: Thompson-Davis v R [2013] NSWCCA 75 at [52].

Section 47(5) focuses on the expiration of the non-parole period of the “other sentence” set by the first court and does not distinguish between the scenarios where the offender is in custody, parole not having been granted, or in custody following the grant of parole and its subsequent revocation: White v R [2016] NSWCCA 190 at [7], [118]–[119]. Therefore, a sentence of imprisonment may not be post-dated later than the earliest date on which the offender will become entitled or eligible to release on parole for the first sentence: White v R at [118]. Basten JA dissented in White v R at [27] on the basis that the:

reference to the offender being “still in custody” [in s 47(5)] is better understood as referring to a continuation of one period of custody rather than the situation where the period of custody has ceased upon his release and recommenced as a result of the revocation of parole.

Where an offender is bail refused for an offence and subject to a statutory parole order pursuant to s 158 Crimes (Administration of Sentences) Act 1999 for a pre-existing sentence, the subject sentence should commence when the non-parole period for the pre-existing sentence expires: Kaderavek v R [2018] NSWCCA 92 at [17]–[22].

[7-550] Information about release date

Last reviewed: August 2023

Section 48(1) Crimes (Sentencing Procedure) Act 1999 provides:

When sentencing an offender to imprisonment for an offence, or to an aggregate sentence of imprisonment for 2 or more offences, a court must specify:


the day on which the sentence commences or is taken to have commenced, and


the earliest day on which it appears (on the basis of the information currently available to the court) that the offender will become entitled to be released from custody, or eligible to be released on parole, having regard to:


that and any other sentence of imprisonment to which the offender is subject, and


the non-parole periods (if any) for that and any other sentence of imprisonment to which the offender is subject.

The three examples given in the Note to s 48(1) are not within the terms of the statute: R v Kay [2000] NSWSC 716. Hulme J said at [128] (affirmed in R v Nilsson [2005] NSWCCA 34):

In specifying the days on which the Prisoner will become eligible for parole and release, I have departed from the examples provided under s 48 of the Crimes (Sentencing Procedure) Act, which reflect a misunderstanding of either simple counting or the law’s measurement of time. Absent special circumstances, the law does not take account of parts of a day. Seven days’ imprisonment commencing on a Monday expires at midnight on the following Sunday.

In Farkas v R [2014] NSWCCA 141, there was a division of opinion as to the appropriate eligibility date of parole. Campbell J at [103] (with whom RA Hulme J agreed at [40]) amended the proposed sentencing orders of Basten JA at [2] so that the applicant’s eligibility for parole fell one day later. Basten JA considered the operation of ss 47 and 48 of the Act, and stated that the parole date which should be specified is that of the day prior to the anniversary of commencement of the sentence: Farkas v R at [29]. His Honour held that there is an inconsistency between the examples set out in the note to s 48 (which assume that the person becomes eligible to be released on parole on the day before the anniversary of the commencement of the sentence) and the language of s 47(6) (“ends at the end of the day on which it expires”). Basten JA opined at [29] that the inconsistency should be resolved by following the approach adopted in the note to s 48 which is consistent with the conventional approach taken in Ingham v R [2014] NSWCCA 123, but see R v Nilsson [2005] NSWCCA 34 at [24], [27]–[29]. While Campbell J or RA Hulme J altered the sentencing orders, neither expressly addressed the operation of s 48.

In R v BA [2014] NSWCCA 148, the court made observations concerning the appropriate date which should be recorded in a parole order. McCallum J stated that the clear effect of s 47(4) is that the Act assumes sentences begin and end at midnight, and it is therefore not inconsistent with the Act to order a person’s release on the last day of the non-parole period. However, such an order could give rise to a technical difficulty in entering the terms of the order into the court’s computerised record system: at [19].

[7-560] Restrictions on term of sentence

Last reviewed: May 2023

Section 49(1) Crimes (Sentencing Procedure) Act 1999 provides:


The term of a sentence of imprisonment (other than an aggregate sentence of imprisonment):


must not be more than the maximum term of imprisonment that may be imposed for the offence, and


must not be less than the shortest term of imprisonment (if any) that must be imposed for the offence.

Section 49(2), which relates to aggregate sentences, is discussed above at [7-505].

[7-570] Court not to make parole orders

Last reviewed: May 2023

Where a non-parole period has been specified for a sentence of 3 years or less, the court must not make an order directing the release of the offender. Section 50 Crimes (Sentencing Procedure) Act 1999, which previously required a court to make such an order, was repealed on 26 February 2018: Parole Legislation Amendment Act 2017, Sch 3.2. However, a court must still comply with s 48(1) Crimes (Sentencing Procedure) Act by nominating when the sentence commences and, when it appears to the court, the offender will be eligible for release: see [7-550] Information about release date.

Section 158 Crimes (Administration of Sentences) Act 1999 states that if a non-parole period has been specified for a sentence of 3 years or less, the offender is taken to be subject to a “statutory parole order”, a parole order directing their release at the end of the non-parole period: s 158(1).

Whenever a court imposes a sentence of imprisonment for a term greater than 3 years, release on parole and the terms of the parole order are matters solely for the Parole Authority: Muldrock v The Queen (2011) 244 CLR 120 at [4]. If the court makes a parole order with conditions in circumstances where it does not have the power to do so “it has no effect”: Moss v R [2011] NSWCCA 86 per Simpson J at [28].

Sections 126 and 158 Crimes (Administration of Sentences) Act are relevant. Section 158(2) provides that a statutory parole order in relation to a sentence is conditional on the offender being eligible for release on parole in accordance with s 126 Crimes (Administration of Sentences) Act at the end of the non-parole period of the sentence. Section 158(3) provides that if the offender is not eligible for release at that time, they are entitled to be released on parole as soon as they become so eligible. Section 158(4) provides that:

This section does not authorise the release on parole of an offender who is also serving a sentence of more than 3 years for which a non-parole period has been set unless the offender is entitled to be released under Division 2.

Section 126 is entitled: “Eligibility for release on parole” and s 126(1) provides that: “Offenders may be released on parole in accordance with this Part”. Section 126(2) provides:

An offender is eligible for release on parole only if:


the offender is subject to at least one sentence for which a non-parole period has been set, and


the offender has served the non-parole period of each such sentence and is not subject to any other sentence.

Mixture of Commonwealth and State offences

In the case of Commonwealth offences, Pt IB Crimes Act 1914 (Cth) makes exhaustive provision for fixing non-parole periods and making recognizance release orders: Hili v The Queen (2010) 242 CLR 520 at [22]. When a court imposes a sentence of 3 years or less (or sentences in aggregate that do not exceed 3 years) on a federal offender, the court must make a recognizance release order in respect of the instant sentence(s) and must not fix a non-parole period: s 19AC(1). The court need not comply with s 19AC(1) if satisfied such an order is not appropriate: s 19AC(4). For further guidance on sentencing, where there is a mixture of Commonwealth and State offences, see Mixture of Commonwealth and State offences at [16-040] Sentencing for multiple offences.

[7-580] No power to impose conditions on parole orders

Last reviewed: August 2023

Following the repeal of ss 51 and 51A Crimes (Sentencing Procedure) Act 1999 on 26 February 2018, the court has no power to impose parole conditions, including conditions as to non-association and place restriction: Sch 3.2[2]–[3] Parole Legislation Amendment Act 2017.

[7-590] Warrant of commitment

Last reviewed: August 2023

As soon as practicable after sentencing an offender to imprisonment, a court must issue a warrant for the committal of the offender to a correctional centre: Crimes (Sentencing Procedure) Act 1999, s 62(1). The warrant must be in the approved form: Crimes (Sentencing Procedure) Regulation 2017, cl 7. Section 62 does not apply to imprisonment the subject of an intensive correction order: s 62(4)(b).

[7-600] Exclusions from Division

Last reviewed: August 2023

Part 4 Div 1 Crimes (Sentencing Procedure) Act 1999 does not apply to offenders sentenced to life (or for any other indeterminate period), or to imprisonment under the Fines Act 1996, the Habitual Criminals Act 1957, or to detention under the Mental Health and Cognitive Impairment Forensic Provisions Act 2020: s 54 Crimes (Sentencing Procedure) Act.