The opinion of the Court was delivered by
O’HERN, J.This appeal concerns the “gap-time credit” provision, found at N.J.S.A 2C:44-5(b)(2) (hereinafter 5(b)(2)). That section governs sentencing at different times for multiple offenses. The provision requires that a defendant who has been sentenced to imprisonment and is subsequently sentenced to another term for an offense committed prior to the imposition of the former sentence (other than an offense committed while in custody) be “credited” at the time of the second sentence for so much of the term of imprisonment as the defendant has served on the prior sentence.
In Richardson v. Nickolopoulos, 110 N.J. 241, 540 A.2d 1246 (1988) (Richardson II), we explained that the use of the expression “gap-time credit” was not to be confused with the usual credit for presentence time referred to as “jail credit.” We explained that 5(b)(2) was a partial adaptation of a provision in the *260Model Penal Code that establishes a “ ‘limit on the cumulation of consecutive sentences.’ ” Id. at 243, 540 A.2d 1246 (quoting Model Penal Code § 7.06 commentary at 272 (1962)). To limit such cumulation, when sentences for multiple offenses are imposed on different occasions, section 7.06 restricts the sentencing authority of courts by requiring them to apply to the later sentence a credit for time served between the two sentencing hearings if the later sentence is for a crime committed prior to the earlier sentence. We explained in Richardson II that “[w]hat the MPC had in mind, then, in speaking of a ‘credit’ was in fact a limit on the maximum term of imprisonment that could be imposed on the sentencing occasion, not a declaration that the time served on the prior sentence be regarded as time served on the current sentence.” Id. at 244, 540 A.2d 1246. The general purpose behind the provision is to avoid the manipulation of trial dates to the disadvantage of defendants and to put defendants in the same position that they would have been “had the two offenses been tried at the same time.” Model Penal Code, supra, § 7.06 commentary at 278.
Our Code does not limit the aggregate of consecutive terms imposed on separate occasions. We thus had to determine in Richardson II what meaning to ascribe to the 5(b)(2) provision when the imposition of the later sentence includes a judicially-imposed parole bar intended to be consecutive to the earlier sentence. Richardson had been sentenced to a five-year term of imprisonment with a two-and-one-half-year parole bar, to be served consecutive to an earlier three-year sentence. Richardson sought to have the two-and-one-half-year parole bar reduced by 336 days of gap time that he had already served on the three-year sentence at the time he received the five-year sentence. We declined to apply gap-time credit to the front end of the second sentence. Otherwise, the credit would reduce the parole bar that the sentencing court had imposed as punishment for the offense. Such an application of the credit would allow the prisoner a free crime because he or she would not serve time for the offense for which the court imposed the second sentence. Rather, we credit gap time to the back end of the sentence. In that way, we *261effectuate the plain language of the statute, which requires time served on the later offense to be credited to the permissible aggregate length of the term or terms remaining to be served. At the same time, we left open in Richardson II the question of how defendants might benefit from our decision to credit gap time to the back end of the sentence.
We granted certification to review two issues posed in these appeals. 134 N.J. 486, 487, 634 A.2d 532 (1993). First, do gap-time credits, as a limit on the total possible sentence, correspondingly reduce the authority of a court to impose a judicial parole bar? Second, do gap-time credits proportionately advance a defendant’s primary parole-eligibility date when neither a judicial nor a statutory parole bar has been imposed? We answer the questions “no” and “yes,” respectively, as did the Appellate Division. Thus, we affirm the judgment of the Appellate Division. However, except for the prisoners currently before us and others already appealing the same issues, we will apply the principles of this decision prospectively in order to minimize administrative difficulties.
I
The cases of Booker and Fitzpatrick present essentially the first question, whether gap-time credits reduce the authority of a court to impose a parole bar; thus, whether we use Booker’s or Fitzpatrick’s case, our analysis and conclusions remain the same. For ease of analysis, we refer only to the facts in Booker. The facts of Fitzpatrick’s case are set out fully in the opinion below. 265 N.J.Super. 191, 194-95, 625 A.2d 1153 (1993). We summarize the facts in Booker as follows. (We eliminate all references to concurrent terms that were imposed at the same sentencing hearing for the sake of simplicity.)
Sentence One for Crime Two (imposed 2(9/90): twenty years with ten-year judicial parole bar.
Sentence Two for Crime One (imposed 5/25/90): fifty years with twenty-five-year judicial parole bar, concurrent to Sentence One.
*262Specifically, Booker asks whether the gap-time credit of 106 days reduces his twenty-five-year parole bar.
First, Booker argues that Richardson II should not apply to him because his situation can be distinguished from that of Richardson on the basis that his sentences are concurrent, whereas Richardson’s were consecutive. He contends that he should be put back where he would have been “had the two offenses been tried at the same time.” Model Penal Code, supra, § 7.06 commentary at 278. His twenty-five-year parole bar, Booker says, should commence to run at the earlier date. That would, in effect, make the sentence retroactive and equate gap time with jail time, a result that we do not believe the Legislature intended. See NJ.S.A 2C:44-5(e)(l) (providing that “[w]hen terms of imprisonment run concurrently, the shorter terms merge in and are satisfied by discharge of the longest term”).
Second, Booker argues that at the very least the 106 days has “perforce” the effect of reducing the second sentence’s parole bar by fifty-three days. He reasons that gap-time credits are a limit on the maximum term of imprisonment that a court can impose at sentencing. Thus, he contends that his second sentence itself has been reduced by 106 days. He argues that because he received the maximum possible judicial parole bar, ie., one-half of his base term of fifty years, he therefore is entitled to have fifty percent of the 106 days (ie., fifty-three days) subtracted from the second sentence’s parole bar of twenty-five years.1 That argument, while internally logical, is not supported by the language of the Code.
N.J.S.A 2C:43-6, which authorizes the imposition of judicial parole bars, does not speak in terms of calculating the parole bar on the basis of the aggregate term that courts may impose. *263Rather, the statute provides that when the aggravating factors substantially outweigh the mitigating factors, the court may “fix a minimum term [of parole ineligibility] not to exceed one-half of the term set pursuant to subsection a. * * N.J.S.A 2C:43-6(b). Subsection (a) establishes the base terms for imprisonment. For example, the base term for a crime of the first degree is between ten years and twenty years; the term for a second-degree crime is between five years and ten years. We are satisfied that the Legislature did not contemplate that the intended effect of 5(b)(2) would be to reduce the authority of courts to impose a ten-year parole bar on a first-degree offense such as aggravated sexual assault.
By way of analogy, we may consider jail credit. If a prisoner had served 365 days before the imposition of a twenty-year sentence for armed robbery, the one-year jail credit would not limit the potential judicial parole bar to nine-and-one-half years (ie., fifty percent of nineteen years). The maximum possible parole bar would still be ten years (ie., fifty percent of twenty years). (Of course, the prisoner would serve only nine years subsequent to sentencing before becoming eligible for parole.) In addition, to assume that the Legislature would expect courts to impose a reduced base term of, for example, nineteen years and 259 days, from which any parole bar would have to be calculated, is unrealistic. Judges do not impose sentences in such terms. The Legislature would not have intended such an administrative irregularity. Hence, we agree with the Appellate Division that “a period of parole disqualifier is an absolute term, against which there are to be no credits (other than jail credits).” 265 N.J.Super. at 207, 625 A.2d 1153. Accordingly, we reject Booker’s contention that his gap-time credit should reduce his parole bar.
II
The question of whether gap-time credits advance primary parole-eligibility dates is much closer. Respondent Darelle Nelson raised the issue. Nelson received the following sentences.
*264Sentence One for Crime Two (imposed 5/11/90): four-year sentence, no parole bar. Sentence Two for Crime One (imposed 12714/90): five-year sentence, no parole bar, concurrent with Sentence One.
Defendant Nelson argues that his gap-time credit of 218 days should advance his primary parole-eligibility date.
In Richardson II, supra, 110 N.J. 241, 540 A.2d 1246, the Attorney General suggested hypothetically that although gap-time credits do not effect a front-end reduction of the aggregate sentence (and thereby reduce any judicial or statutory parole bar), they may effect a back-end reduction of the aggregate sentence and thus advance proportionately the primary parole-eligibility date, which is a function of the aggregate sentence. Under that interpretation, the effect of the provision is to reduce the total sentence by the gap period, thus requiring the Parole Board to calculate primary parole eligibility based on the new reduced sentence, which would result in a shorter parole-ineligibility period. Richardson had argued that unless we applied his gap-time credits to his parole bar, the statute would be meaningless. The interpretation now adopted by defendant Nelson was suggested merely as a means of responding to Richardson. Because the issue was not properly before us in Richardson II, we did not resolve it.
The Parole Board urges us not to accept Nelson’s argument. N.J.S.A 30:4-123.51(a) (the Parole Act) provides, in part, the following:
Each adult inmate sentenced to a term of incarceration in a county penal institution, or to a specific term of years at the State Prison or the correctional institution for women shall become primarily eligible for parole after having served * * * one-third of the sentence imposed where no mandatory minimum term has been imposed less commutation time for good behavior * * * and credits for diligent application to work and other institutional assignments * * *.
The Parole Board emphasizes the plain language of the statute, “one-third of the sentence imposed,” and explains that it is only when the sentencing court itself reduces the sentence that parole eligibility is advanced. The Board reasons that the Legislature would have mentioned that gap-time credits reduce the one-third period if that is what the Legislature had intended. We are *265informed, however, that jail credits, although also not mentioned in the Parole Act, do serve to advance primary parole-eligibility dates. The question is close, but we believe that if we are to give any meaning at all to the provision, it must involve some reduction of the cumulative period of time to be served—a limit on the maximum term of imprisonment that can be imposed at sentencing. The parole laws provide that prisoners will become primarily eligible for parole when they have served one third of their sentences. As with other types of sentencing credits, gap-time credits must be determined by the court at sentencing. The Parole Board is not responsible for awarding such credits. We could remand such matters to the trial courts for recalculation of the sentences imposed, but we think that such a solution would be unwieldy. Because the other administrative credits must be calculated by the Parole Board (and because those credits may change for misconduct), it makes sense that once credits have been awarded by the sentencing court, “the Parole Board must compute defendant’s parole eligibility date on the basis of the reduced aggregate sentence.” 265 N.J.Super. at 208, 625 A.2d 1153. We do not anticipate the same administrative problems that would have arisen in attempting to apply gap-time credits to alter judicial parole bars. In essence, the prisoner receives for the gap time a one-third credit on the actual time to be served before parole eligibility commences.
Ill
A final question is whether gap-time credits affect the aggregate of consecutive sentences. For example, assume that Nelson’s sentences had been consecutive rather than concurrent. Would the 72 days of credit (one-third of the 218 days of gap time) reduce the actual time to be served on the nine-year aggregate term? The opinion below suggests that in the case of consecutive sentences the credit does not advance primary parole-eligibility dates because the purposes of the credit are not implicated. Returning to the Model Penal Code, the purpose of gap-time credits is to put the defendant in the position in which he would *266have been “had the two offenses been tried at the same time.” Model Penal Code, supra, § 7.06 commentary at 278. Whether Nelson receives five years consecutive to four or four years consecutive to five, so the argument goes, makes no difference. In either case the aggregate is the same, nine years, unlike the situation in which a delay in sentencing on an earlier five-year term to be served concurrent to a four-year term can result in more jail time. That argument, like Booker’s argument, has logical appeal, but the plain language of the gap-time statute clearly states that the credits apply “[wjhether the court determines that the terms shall run concurrently or consecutively.” N.J.S.A 2C:44-5(b)(2) (emphasis added). Circumstances may arise in which the interplay of jail credits (given only for the offense for which the prisoner is serving time) and gap-time credits (given for other sentences being served by the prisoner) could affect the total time to be served. Furthermore, an across-the-board application of gap-time credits to both concurrent and consecutive sentences achieves the provision’s purpose to deter delay (whether due to dilatory tactics or simple unavailability of court or counsel). After all, the prosecutor will not know in advance if the sentences are to be consecutive or concurrent. Hence, we hold that the gap-time statute applies to reduce the aggregate of consecutive sentences.
Our holding does not contradict the Attorney General’s Advice Letter AAA No. M85-6515 or the Administrative Office of the Court’s memorandum dated October 15, 1985, both of which declare that the determination of gap-time credits is part of the sentencing process to be conducted by the court. We foresee no administrative difficulty in the Parole Board molding its regulations to reflect that the “sentence imposed,” N.J.SA 30:4-123.-51(a), should be reduced by any gap-time credits that the prisoner has been awarded.
Finally, we are aware of the problem that a retroactive ruling would create. Retroactivity would force the Parole Board to recalculate primary parole-eligibility dates in the many cases in *267which gap-time credits have been awarded. The Parole Board has undoubtedly set its internal procedures in place for that very important final decision on parole. Given that 5(b)(2) has been codified since 1979, applying this ruling prospectively, except for cases now on appeal, can hardly be thought to raise an ex post facto problem.
IV
We are cognizant of the Act’s internal inconsistencies as well as the substantial difficulties in applying the Act. “At the present time, gap-time credits remain ‘a riddle wrapped in a mystery inside an enigma.’ ” State v. Edwards, 263 N.J.Super. 256, 262, 622 A.2d 919 (App.Div.1993). As counsel for appellant Booker puts it, “Although the Soviet Union has disintegrated, the question of ‘gap time’ remains.” We wish that we could write a computer program that would answer each gap-time question as it arises. We believe, however, that the Appellate Division’s disposition achieves the main purpose of gap-time credits.
In summary, we hold that where a defendant is sentenced to a period of parole ineligibility, gap-time credit only reduces the aggregate term to be served after the defendant serves the mandatory period of incarceration. Where a defendant receives no mandatory period of incarceration, gap-time credit reduces the aggregate term and the Parole Board must compute defendant’s parole eligibility date on the basis of the reduced aggregate sentence.
[265 N.J.Super. at 208, 625 A.2d 1153.]
If gap-time credits do not affect prisoners’ parole-eligibility dates, we would truly have what the Public Defender calls “bogus credits,” a calculation without any effect. Subtracting 106 days from the back end of a fifty-year sentence does not confer any considerable benefit on most offenders because those offenders with long parole bars will have “maxed out” long before the credit becomes relevant. On the other hand, in cases of relatively short sentences involving a fairly substantial period of gap time, as for example in Nelson’s case, the credits serve a valid legislative purpose.
*268Our concurring and dissenting member makes a persuasive argument against our resolution of the issues. He is correct in asserting that crediting gap time to the front end of concurrent sentences would better achieve the goal of putting defendants in the same position in which they would have been “had the two offenses been tried at the same time.” Model Penal Code, supra, § 7.06 commentary at 278. However, at the same time, that interpretation would lead in many cases to the neutralization of mandatory-minimum periods of parole ineligibility, a result that we believe the Legislature never intended. Hence we adhere to our holding in Richardson II, supra, 110 N.J. 241, 540 A.2d 1246, that gap-time credits apply to the back end of sentences. That holding applies to consecutive sentences as well as concurrent sentences, as indicated by the plain language of the statute. Application of the credits to consecutive sentences comports with our basic understanding that the credits are a “‘limit oh the cumulation of consecutive sentences.’ ” Id. at 243, 540 A.2d 1246 (quoting Model Penal Code, supra, § 7.06 commentary at 272).
On a final note, candor requires us to reiterate Judge Baime’s recent observation about this “arcane subject”: “Perhaps it is time for us to admit defeat and tell the Legislature in plain terms that we do not know what that body intended when it enacted the gap time statute. With our respect, we commend this matter to the Legislature for its consideration.” State v. Guaman, 271 N.J.Super. 130, 135, 638 A.2d 162 (App.Div.1994).
The judgments of the Appellate Division are affirmed.
The argument appears rather insignificant in the context of Booker's sentence of fifty years, which is a long time to wait for a 106-day credit. However, this case is atypical. Booker's base term on Sentence Two is so great because he pleaded guilty to five multi-count indictments and agreed to the aggregate fifty-year term. Others with much shorter sentences could benefit substantially if we were to adopt Booker's position.