On 9 October 2026, a French extradition case in London collapsed without a judge ever ruling on it. The Crown Prosecution Service informed the High Court that the National Crime Agency had been told the French arrest warrant was withdrawn. The requested person, wanted by France to serve a two-year prison sentence imposed for offences dating from 2014-2015 after a conviction in his absence, had been arrested in the United Kingdom in August 2024. Extradition had been ordered in April 2025. Permission to appeal was refused on the papers in March 2026. Everything pointed to a routine surrender. Then the arithmetic of time changed the outcome: seven days in custody plus an uninterrupted electronically monitored curfew running every night from 7pm to 6am since 31 August 2024. According to an updating opinion from Paris Bar lawyer Paul Aprile, acting for the requested person in France, that period was deductible from the French custodial sentence under French law. By September 2026, the custody and curfew credit had exceeded the two-year sentence, leaving no custodial balance to serve. The defence invited the court to ask why a man should be removed from his established private and family life to serve a sentence already exhausted. Before the renewed permission hearing listed for 13 October 2026 could take place, the warrant was withdrawn.
This development, reported on 9 October 2026 by extradition barrister George Hepburne-Scott of Church Court Chambers, who represented the requested person, matters far beyond one file. It exposes the exact mechanism by which French law can neutralise a French extradition request from the inside: the imputation of time spent under foreign restrictions on the French sentence. And it lands only four months after the Court of Cassation drew a bright line on that very question. On 10 June 2026, the criminal chamber rejected a convicted person’s attempt to deduct a British curfew from an eighteen-month French sentence, holding that a midnight-to-5am curfew was not equivalent to a French assignation à résidence sous surveillance électronique. Yet on 17 March 2021, the same chamber upheld the deduction of a 10pm-to-5am British curfew with a tag and daily police reporting. Between five hours a night, which fails, and the eleven-hour curfew in the October 2026 London case, which helped kill the warrant, sits one of the most practical defences available to anyone wanted by France and arrested abroad. This article explains, from the French-law side, how the curfew-credit argument works, where courts draw the line, and how the wider French extradition machinery — the investigating chamber’s opinion, the Prime Minister’s decree, limitation rules, and the remedies after surrender is ordered — frames every such fight.
I. How a British Curfew Can Wipe Out a French Prison Sentence
A. The 9 October 2026 London Breakthrough That Killed a French Warrant
The facts reported on 9 October 2026 are worth setting out precisely, because each detail fed the legal argument. The French request concerned a two-year prison sentence imposed following a conviction in absence for four offences connected with alleged involvement dating from 2014-2015. The client was arrested in the United Kingdom in August 2024, and extradition was ordered in April 2025. The defence challenged the order in the High Court. Permission to appeal was refused on the papers in March 2026, but the defence continued through a renewed application supported by fresh evidence, centred on Article 8 of the European Convention on Human Rights and the proportionality of extradition. That persistence proved decisive, because the case developed beyond the circumstances that existed when extradition was originally ordered: as time passed, the legal significance of the lengthy curfew grew until it consumed the sentence itself.
The fresh-evidence application relied on an updating opinion from Paul Aprile, a lawyer at the Paris Bar acting for the client in France. The evidence recorded seven days in custody and an uninterrupted electronically monitored curfew beginning on 31 August 2024, running daily from 7pm to 6am. Mr Aprile’s opinion was that, under the applicable French law, the relevant period was deductible from the French custodial sentence. His conclusion was unequivocal: by September 2026, the custody and curfew credit had exceeded the two-year sentence, leaving no custodial balance to serve. The defence submitted that removing the client from his established private and family life to enforce a sentence already exhausted would fundamentally alter the Article 8 balance, and French counsel had repeatedly approached the prosecuting authorities seeking withdrawal of the warrant or consideration of an alternative mechanism. Following the defence request, the CPS made enquiries of the French authorities, and its subsequent update reported the NCA’s information that the warrant had been withdrawn. The CPS email does not give the French authorities’ reasons for withdrawal, and no High Court determination accepting the defence arguments has yet been made — authenticated confirmation and the formal court steps were still awaited at the time of the report. The lesson is nonetheless clear, and it is a French-law lesson: the sentence described in a warrant must be reassessed against custody, bail restrictions and the law of the requesting country, because a curfew, combined with the right foreign-law evidence, can exhaust the very sentence extradition is supposed to enforce.
The application was built on a line of English authorities concerning the significance of United Kingdom electronically monitored curfews in French extradition cases, including A v France [2022] EWHC 3214 (Admin), Doga v France [2023] EWHC 2561 (Admin) and Prisacaru v France [2025] EWHC 1416 (Admin). Readers who want the full French procedural map — the investigating chamber’s opinion, appeals, the decree, and the guarantees of the person sought — will find it in our complete guide to French extradition procedure before the chambre de l’instruction. The remainder of this first part stays on the narrower and more explosive question: when does French law deduct foreign curfew time from a French sentence, and where exactly is the boundary?
B. The French Deduction Machinery: What Courts Deduct and Where They Stop
French law starts from a generous principle. Article 716-4 of the Code of Criminal Procedure provides that “Quand il y a eu détention provisoire à quelque stade que ce soit de la procédure, cette détention est intégralement déduite de la durée de la peine prononcée”. The same article extends the rule beyond French soil: “Il en est de même, s’agissant d’une détention provisoire ordonnée dans le cadre d’une procédure suivie pour les mêmes faits que ceux ayant donné lieu à condamnation, si cette procédure a été ultérieurement annulée”, and the provisions are “également applicables à la privation de liberté subie en exécution d’un mandat d’amener ou d’arrêt, à l’incarcération subie hors de France en exécution d’un mandat d’arrêt européen ou sur la demande d’extradition”. In other words, time spent locked up abroad on a European Arrest Warrant or an extradition request counts, in full, against the French sentence. The bridge to curfews is Article 142-11 of the Code of Criminal Procedure: “L’assignation à résidence avec surveillance électronique est assimilée à une détention provisoire pour l’imputation intégrale de sa durée sur celle d’une peine privative de liberté”. And the underlying measure, defined by Article 142-5 of the Code of Criminal Procedure, “peut être ordonnée, d’office ou à la demande de l’intéressé, par le juge d’instruction ou par le juge des libertés et de la détention si la personne mise en examen encourt une peine d’emprisonnement correctionnel d’au moins deux ans ou une peine plus grave”, with the statute adding that “Cette obligation est exécutée sous le régime du placement sous surveillance électronique”. The defence syllogism in every curfew-credit case is therefore simple: a British electronically monitored curfew is the functional twin of a French assignation à résidence sous surveillance électronique; that twin is assimilated to pre-trial detention; pre-trial detention is fully deducted, including when suffered abroad on a European warrant or an extradition request. The fight is always about the first premise — equivalence — and French courts answer it with a four-factor test inherited from Luxembourg.
The test comes from the Court of Justice of the European Union’s judgment of 28 July 2016 in JZ (C-294/16), as summarised by the French courts: a nine-hour night-time house arrest with an electronic tag, combined with regular police reporting and a ban on seeking travel documents, does not in principle constitute detention within the meaning of Article 26(1) of Framework Decision 2002/584/JHA. Whether a foreign measure counts as detention depends on an overall assessment of its type, duration, effects and detailed enforcement arrangements, and only measures whose constraining effect is comparable to incarceration qualify. French courts apply that grid twice: once under European Union law, once under domestic law, because Article 26 of the Framework Decision only sets a minimum floor and French law may be more generous through Articles 142-11 and 716-4.
The two leading French rulings point in opposite directions, and the contrast is the whole strategy manual. In the first, the Court of Cassation, criminal chamber, 17 March 2021, No. 20-84.365, two men arrested in the United Kingdom on 26 February 2018 on Paris European Arrest Warrants spent thirty-two days in detention and then, from 29 March 2018 until their surrender on 23 May 2019, lived under “bail with curfew electronically monitored”: curfew at their residence from 10pm to 5am under electronic control, a ban on going to certain places, daily reporting to the police station, a permanently worn ankle tag, and a telephone switched on at all times. Sentenced in Paris to three years’ imprisonment, they asked for the whole British period to be deducted. The Paris court agreed, and the public prosecutor appealed. The Court of Cassation rejected the appeal, endorsing the finding that “il y a lieu de l’assimiler, en droit français, à une mesure d’assignation à résidence sous surveillance électronique qui n’opère aucune distinction selon la durée quotidienne de l’obligation de rester au domicile”. A seven-hour nightly curfew, combined with a tag, a phone obligation and daily reporting, counted as a French-style electronically monitored house arrest, and its duration was fully deducted. In the second ruling, the Court of Cassation, criminal chamber, 10 June 2026, No. 25-84.351, a man sentenced by the Lyon Court of Appeal to eighteen months’ imprisonment, subject to a European Arrest Warrant, was arrested in the United Kingdom on 16 November 2020 and placed the next day under electronically monitored curfew until 3 October 2023, when his extradition was refused. He asked the Lyon court to lift the arrest warrant and declare the sentence served. The court refused, and the Court of Cassation upheld the refusal: the judges had been entitled to find that the measure, “eu égard à sa nature, à sa durée, à ses effets et à ses modalités d’exécution”, was not detention within the meaning of Article 26(1) of the Framework Decision, and they had then correctly asked, as they had to, whether under French law the British measure could be assimilated to the electronically monitored house arrest defined by Article 142-5. Their answer was no: the man had been subjected for thirty-four months and sixteen days to a restriction of his freedom of movement, and “le seul fait qu’elle était contrôlée par un dispositif électronique ne justifie pas de la regarder comme équivalente à une assignation à résidence sous surveillance électronique dès lors qu’il gardait pour l’essentiel, chaque jour, pendant dix-neuf heures, toute latitude de se rendre, pour quelque motif que ce soit, où bon lui semblait”. A five-hour curfew, from midnight to 5am, even with a permanent telephone obligation, twice-weekly police reporting and a ban on leaving the United Kingdom, was a restriction of movement, not a deprivation of liberty, and nothing was deducted.
Read together, the two rulings give practitioners the working grid. A short night-time curfew with few accessory obligations fails; a longer curfew with a tag, constant telephone availability and frequent reporting succeeds. The October 2026 London case sat well above the successful precedent: eleven hours a night, from 7pm to 6am, uninterrupted since 31 August 2024, plus seven days of custody, exceeding a two-year sentence. That is why the French-law opinion could state unequivocally that nothing remained to be served, and why the requesting authorities, once seised through the CPS enquiry, chose withdrawal over a hearing. Three practical consequences follow. First, the equivalence analysis is sovereign: the Court of Cassation reviews only whether the lower court applied the right test, so the battle is won or lost on evidence — daily schedules, tag records, reporting logs, telephone obligations, travel bans — placed before the court seized of the incident. Second, the request must be framed as a difficulty in enforcement before the French court, with a precise day-by-day count, because only documented constraint converts into deducted days. Third, time is the defence’s ally in a way rarely acknowledged: every additional month under curfew while the appeal proceeds increases the credit and shrinks the sentence left to enforce, progressively unbalancing the Article 8 proportionality assessment until, as in October 2026, the justification for removal collapses entirely.
II. How France Decides Whether to Surrender You or Prosecute You Instead
A. Who Gets Surrendered, Who Gets Tried in France, and When Time Bars the Request
Curfew credit decides whether a sentence remains to be served. A logically prior question decides whether France may surrender anyone at all: extradition is refused whenever the statutory bars are met. Article 696-4 of the Code of Criminal Procedure lists them, and the list is worth reading in full: “L’extradition n’est pas accordée : 1° Lorsque la personne réclamée a la nationalité française”, with nationality assessed at the time of the offence; for political offences or requests made for a political purpose; for offences committed on French territory; for offences already finally tried in France; and, crucially for sentence-enforcement cases, “Lorsque, d’après la loi de l’Etat requérant ou la loi française, la prescription de l’action s’est trouvée acquise antérieurement à la demande d’extradition, ou la prescription de la peine antérieurement à l’arrestation de la personne réclamée et d’une façon générale toutes les fois que l’action publique de l’Etat requérant est éteinte”. Two further bars close the article: punishment by the requesting state with a penalty contrary to French public policy, trial by a court not guaranteeing fundamental procedural safeguards and defence rights, and purely military offences.
Two recent rulings show how the limitation bar is litigated in practice. The Court of Cassation, criminal chamber, 28 January 2026, No. 25-81.433, ruled on a Turkish request concerning cumulative sentences, including a 162-day prison term converted into a fine and then contested as time-barred. The Court restated the governing rule: “d’une part, l’extradition n’est pas accordée lorsque la prescription de l’action ou de la peine est acquise d’après la législation de la partie requérante, et n’est pas refusée au motif qu’elle serait acquise d’après la législation française, sauf lorsque la demande d’extradition se fonde sur des infractions pour lesquelles les juridictions françaises seraient compétentes en vertu de la loi française, d’autre part, la prescription de la peine s’apprécie à la date de l’arrestation de la personne réclamée”. The Versailles investigating chamber had computed limitation from the pronouncement of the fine; the Turkish authorities indicated that time ran from the conversion notification of 1 April 2015, pushing the ten-year expiry to 1 April 2025, after the arrest. The Court quashed the chamber’s finding: “l’éventuelle acquisition de la prescription relevant, hors le cas où l’infraction pourrait être poursuivie en France, des règles qui la gouvernent dans l’Etat requérant, telles qu’elles résultent des informations fournies par ce dernier”. The same decision carries a procedural warning that defence teams ignore at their peril: the requested person’s own appeal was declared inadmissible because it had been filed by a Paris Bar lawyer who neither practised before the deciding court nor produced the special authority required, since the appeal “est formé par un avocat qui, en premier lieu, n’exerce pas près la juridiction ayant statué, et en second lieu, n’a pas justifié du pouvoir spécial exigé par l’article 576 du code de procédure pénale”. For limitation periods themselves, the scales are fixed by the Criminal Code: Article 133-3 of the Criminal Code provides that “Les peines prononcées pour un délit se prescrivent par six années révolues à compter de la date à laquelle la décision de condamnation est devenue définitive”, while the companion articles set twenty years for felonies and three years for petty offences. Note carefully what the 28 January 2026 ruling actually decides: it fixes the assessment date at the arrest and designates the applicable law. It does not create any suspension, interruption, postponement or starting-point rule of its own, and no such judge-made extension should be read into it.
Where extradition is refused, France does not necessarily walk away. A French national cannot be extradited at all, and a settled line of cases allows France to prosecute in France facts committed abroad once surrender is off the table. That is the second half of the forum choice every requested person must make with counsel: resisting surrender may mean facing a French court instead, under French penalties and French limitation rules, with the advantage of a home forum and the drawback of a fresh prosecution. Convictions in absence add a further layer. The October 2026 London case involved a conviction in absence, and French law treats default judgments in correctional matters as fragile by design: Article 489 of the Code of Criminal Procedure states that “Le jugement par défaut est non avenu dans toutes ses dispositions, si le prévenu forme opposition à son exécution”. Anyone surrendered to France on a default conviction therefore holds a statutory second chance — opposition wipes the judgment out and the case is retried with the defendant present — which is precisely why requesting states sometimes prefer to proceed on a warrant that the defence can attack as already served.
B. Fighting the Request From Paris: Custody, Control, Opinion and Decree
Procedure is where extradition cases are won, and the French track is codified step by step. Article 696-11 of the Code of Criminal Procedure governs the first hours: after notification of the request, if the public prosecutor decides not to leave the requested person at liberty, the person is brought before the first president of the court of appeal, who orders committal under an extradition warrant — but “s’il estime que sa représentation à tous les actes de la procédure est suffisamment garantie, le premier président de la cour d’appel ou le magistrat du siège désigné par lui peut soumettre la personne réclamée, jusqu’à sa comparution devant la chambre de l’instruction, à une ou plusieurs des mesures prévues aux articles 138 et 142-5”, meaning judicial supervision or electronically monitored house arrest instead of prison. Article 696-15 of the Code of Criminal Procedure then sets the core hearing: “Lorsque la personne réclamée a déclaré au procureur général ne pas consentir à son extradition, la chambre de l’instruction est saisie, sans délai, de la procédure”, the person appears within ten working days, the chamber delivers its reasoned opinion within one month, and “Cet avis est défavorable si la cour estime que les conditions légales ne sont pas remplies ou qu’il y a une erreur évidente”. The appeal is deliberately narrow: “Le pourvoi formé contre un avis de la chambre de l’instruction ne peut être fondé que sur des vices de forme de nature à priver cet avis des conditions essentielles de son existence légale.” Liberty during the proceedings is protected by Article 696-19 of the Code of Criminal Procedure: “La mise en liberté peut être demandée à tout moment à la chambre de l’instruction”. And the end of the road, when the opinion is favourable, is political: Article 696-18 of the Code of Criminal Procedure provides that, outside the case covered by Article 696-17, “l’extradition est autorisée par décret du Premier ministre pris sur le rapport du ministre de la justice”, with a strict one-month time limit on pain of foreclosure: “Le recours pour excès de pouvoir contre le décret mentionné à l’alinéa précédent doit, à peine de forclusion, être formé dans le délai d’un mois.” Every deadline in this paragraph comes from the cited statute; the article asserts no other time limit of its own.
Two rulings complete the practitioner’s picture of the after-opinion phase. The Court of Cassation, criminal chamber, 9 August 2023, No. 23-83.192, decided that a person under judicial supervision does not automatically walk free when the investigating chamber issues an unfavourable opinion, if the public prosecutor appeals. The chamber had held that under Article 696-17 the requested person is released as of right only where the opinion rejecting extradition is final, adding that “la Cour de cassation étant saisie de l’avis défavorable par le pourvoi du procureur général, ladite décision n’est pas définitive”. The Court of Cassation approved: since an appeal suspends execution, “il s’induit des dispositions de l’article 696-17 du même code que la mainlevée de la mesure de contrôle judiciaire dont fait l’objet la personne réclamée ne saurait être acquise dès le prononcé d’un arrêt de la chambre de l’instruction qui a émis un avis défavorable à son extradition”. In plain terms, a favourable outcome at the investigating chamber can still leave the person under supervision for months while the prosecutor’s appeal runs its course — a period to use, not to endure passively. The operational checklist for a Paris defence team therefore writes itself. First, commission the French-law opinion early: as the October 2026 case shows, a reasoned Paris Bar analysis of deductibility, served on the requesting authorities through the executing-state prosecution, can produce withdrawal before any court rules. Second, build the equivalence file from day one — curfew hours, tag data, reporting records, telephone and travel obligations — because the 2021 and 2026 rulings turn entirely on documented constraint, and the Court of Cassation will not second-guess the lower court’s sovereign assessment of it. Third, run the limitation audit in parallel under the law of the requesting state as communicated by that state, measured at the date of arrest, and plead opposition readiness if the file rests on a default conviction. Fourth, calendar the decree: if the opinion is favourable and a Prime Minister’s decree follows, the one-month judicial-review deadline is foreclosure, and the grace petition does not stop the clock. Handled this way, an extradition request becomes what the October 2026 withdrawal proves it can be — not a conveyor belt to surrender, but a file that French law itself can empty.
Conclusion
The withdrawal reported on 9 October 2026 is the clearest recent proof that extradition to France is a contest of French law, not a formality. A two-year sentence for 2014-2015 offences, a conviction in absence, an arrest in August 2024 and an extradition order in April 2025 looked like the end of the story. Eleven-hour electronically monitored nights, documented without interruption since 31 August 2024 and credited under Articles 142-11 and 716-4 of the Code of Criminal Procedure, rewrote it. The Court of Cassation’s twin rulings frame the outcome: the 17 March 2021 decision upheld deduction of a seven-hour tagged curfew with daily reporting, while the 10 June 2026 decision refused deduction for a five-hour curfew that left nineteen hours of daily freedom, each time by asking whether the foreign measure, in type, duration, effects and arrangements, matched a French electronically monitored house arrest. Around that core, the surrounding machinery — refusal bars and limitation measured at arrest under the requesting state’s law, opposition wiping out default convictions, release applications at any time, supervision surviving an unfavourable opinion under appeal, and the Prime Minister’s decree challengeable within one month on pain of foreclosure — gives every requested person a structured set of moves. The side that masters the count of days, the paper trail of constraint and the calendar of remedies is the side that decides whether the warrant ends at a hearing, at a decree, or, as in London this month, in a quiet withdrawal before the hearing ever happens.