
As my beloved passes her first month in prison, I have met a journalist who is keeping a tab on the Turkish prison system here (Prison Agenda). I have asked ChatGPT to prepare an in-depth report on the Turkish prisons. As usual, the government has been proud of building bigger prisons in recent years, but even this does not reach the level of imprisonment, and prisons remain overcrowded.
Behind the Walls: Türkiye’s Prison System at a Record Scale in 2026
Executive summary
Türkiye entered November 2026 with 437,889 people in prison, according to the General Directorate of Prisons and Detention Houses (Ceza ve Tevkifevleri Genel Müdürlüğü, CTE): 365,758 sentenced prisoners and 72,131 remand detainees. The remand share was therefore about 16.5% of the total. The same dataset counted 21,524 adult women and 5,121 children, including 3,809 children held on remand. [1]
That headline number represents an extraordinary transformation. The prison population was roughly 49,500 in 2000, about 120,800 in 2010 and 200,000 in 2016. By the end of 2025 it had reached 401,519, before climbing another 36,000-plus during the first ten months of 2026. [2] In other words, Türkiye’s prison population is now almost nine times its 2000 level. Using the latest complete population denominator available from TÜİK—86.09 million residents at the end of 2025—the November 2026 prison stock would correspond to roughly 509 prisoners per 100,000 inhabitants; this is only an indicative rate because the numerator and denominator refer to different dates. [3]
The last fully harmonised European comparison is less current but methodologically much stronger. In the Council of Europe’s SPACE I survey for 31 January 2025, Türkiye held 392,456 prisoners in an officially reported operational capacity of 299,880—130.9 prisoners per 100 places—and had an imprisonment rate of 458.1 per 100,000. The Council of Europe identified Türkiye as having Europe’s highest imprisonment rate and, together with France, one of its most severely overcrowded systems among large member states. The median European imprisonment rate was around 110–115 per 100,000, depending on the comparison set. [4]
The central finding of this investigation is therefore not merely that Turkish prisons are overcrowded. It is that penal expansion has become structural. Capacity building has repeatedly lagged behind population growth; temporary release and probation reforms have produced sharp short-term reductions without altering the long-term upward trajectory; high-security institutions impose regimes that the European Committee for the Prevention of Torture (CPT) considers excessively isolating; and weaknesses in judicial independence, pretrial detention practice and broad terrorism legislation feed into imprisonment patterns that cannot be understood as a prison-administration problem alone. [5]
At the same time, an evidence-based account must distinguish established findings from allegations. The CPT found generally satisfactory material conditions in several high-security prisons visited in 2024 and received no recent allegations of physical ill-treatment at three of four such institutions; at Antalya S-type Prison, however, it recorded several allegations it judged credible, including one case where recent injuries were consistent with the prisoner’s account. Turkish authorities reject claims of systematic abuse, cite a “zero tolerance” policy toward torture, point to investigations and disciplinary measures, investments in new institutions, expanded health services and extensive education and work programmes, and argue that overcrowding is managed operationally through additional beds and transfers. [6]
The picture that emerges is therefore neither one of a uniformly abusive prison estate nor one reducible to individual high-profile political cases. It is a mass incarceration system under extreme quantitative pressure, containing substantial rehabilitative infrastructure but suffering from serious capacity, isolation, medical-care, monitoring and rule-of-law problems. The most credible international bodies increasingly argue that building more prisons cannot by itself solve those problems. [7]
Key figures
| Indicator | Date | Figure | Evidence note |
| Total prison population | 2 Nov 2026 | 437,889 | CTE administrative data; 365,758 sentenced, 72,131 remand. [8] |
| Remand detainees | 2 Nov 2026 | 72,131 / 16.5% | Calculated from CTE stock figures. [8] |
| Children in prison | 2 Nov 2026 | 5,121 | 1,312 sentenced; 3,809 remand. [9] |
| Adult women in prison | 2 Nov 2026 | 21,524 | CTE. [9] |
| Prison population | 31 Dec 2025 | 401,519 | Official annual prison statistics. [10] |
| Official capacity | 31 Dec 2025 | 305,086 | Official annual prison statistics; capacity and population share the same reference date. [10] |
| SPACE prison population | 31 Jan 2025 | 392,456 | Harmonised Council of Europe comparison. [11] |
| SPACE capacity | 31 Jan 2025 | 299,880 | “Operational capacity”; Türkiye notes that extra beds can be added. [11] |
| SPACE prison density | 31 Jan 2025 | 130.9 per 100 places | Severe overcrowding by Council of Europe classification. [4] |
| SPACE imprisonment rate | 31 Jan 2025 | 458.1 per 100,000 | Highest among Council of Europe systems. [4] |
| Women | 31 Dec 2025 | 19,068 / 4.7% | Annual official statistics. [12] |
| Foreign nationals | 31 Dec 2025 | 15,759 / 3.9% | Annual official statistics. [12] |
| Prison staff | 31 Dec 2025 | 77,737 | Official annual statistics; 5.2 prisoners per employee. [10] |
| Deaths reported in prisons | 2025 | 1,001 | Figure stated in Parliament: 869 described as natural deaths and 132 other deaths, including accidents/leave-related deaths. [13] |
| Prison-work participation | 2025 | 77,186 during the year | İşyurtları annual report. [14] |
| Prison-work participation | through 30 Oct 2026 | about 82,000 during 2026 | Justice Ministry statement; more than 32,000 participating at that moment. [15] |
A methodological warning is essential. The November 2026 prisoner count should not simply be divided by the December 2025 capacity figure and presented as an official 2026 occupancy rate, because new prisons and capacity changes during 2026 may alter the denominator. The closest harmonised European density measurement remains SPACE I’s January 2025 figure of 130.9%. [16]
Prison Agenda: what it is, what it adds, and how to read it critically
The English-language Prison Agenda Substack describes itself as an “independent, rights-based digital newsletter” focused on Türkiye’s prison ecosystem, explicitly combining data analysis with a human-rights perspective. Its public “About” page says its objective is to make prisons more visible through data-driven, analytical reporting rather than to “romanticize” imprisonment. [17]
The publication was founded by Ahmet Kavruk. Its first English issue identifies Kavruk and explains that he and Tutku Sönmez had been producing the project together since March 2026, after a Turkish-language version was launched earlier in the year. Kavruk says he had worked on prisons and prisoners’ rights for more than a decade in professional and voluntary capacities. His own professional webpage likewise describes Prison Agenda/Gündem Hapishane as an independent digital platform that he established in early 2026 and through which he conducts investigative journalism, data analysis and visualisation. [18]
Its editorial architecture is unusually useful for outsiders because it turns several otherwise scattered sources into recurring products: a monthly news bulletin, a quarterly statistical bulletin using CTE data, parliamentary monitoring, and thematic articles on women, children, LGBTI+ prisoners, international standards and torture prevention. The publication states that it received support under the Digital Spark grant scheme associated with Metamorphosis Foundation, TechSoup, CIVICUS and the Digital Democracy Initiative. [19]
Its strongest work is where it functions as data translation rather than primary investigation. Its July 2026 article on SPACE I, for example, correctly highlighted Türkiye’s 392,456 prisoners, 299,880 operational capacity, density of 130.9 per 100 places and imprisonment rate of roughly 458 per 100,000. Those core claims are independently confirmed by the Council of Europe/University of Lausanne data. [20] Its monthly bulletins likewise reproduce CTE stock figures accurately: the October bulletin reported 434,337 prisoners as of 1 October 2026, including 71,799 on remand; the official CTE series subsequently reached 437,889 by 2 November. [21]
Other especially relevant Prison Agenda pieces include explainers on the Nelson Mandela Rules, girls in prison, LGBTI+ prisoners and the UN Convention against Torture. Those pieces serve a different function: they interpret Turkish practice through international human-rights standards rather than merely describing administrative data. [22]
That orientation is both a strength and a perspective that readers should recognize. Prison Agenda is openly rights-centred, not normatively neutral; it approaches imprisonment primarily through dignity, transparency and vulnerability rather than deterrence or punitive policy. That does not invalidate its factual work, but it does mean evaluative language should be checked against primary evidence. Its comparative article, for instance, described Türkiye as among Europe’s most overcrowded systems, which is supported by SPACE; however, its wording that Europe’s “average” density stood at 86 differs from the Council of Europe’s headline continental figure of 95.2 prisoners per 100 places, reflecting the importance of distinguishing medians, averages, reporting subsets and exclusion rules. [23]
The publication should therefore be treated as a high-value specialist secondary source: particularly strong for finding official statistics, monitoring legislation and identifying underreported groups, but not as a substitute for CTE, TÜİK, SPACE, CPT, ECtHR or UN documentation when making contested claims. That distinction is consistent with the founders’ own explanation that their purpose is largely to transform scattered and technical public information into accessible reporting. [24]
From incarceration growth to chronic overcrowding
Türkiye’s prison expansion long predates the post-2016 political crisis. Council of Europe historical data show the imprisonment rate rising from about 76 per 100,000 in 2005 to 220 per 100,000 in 2015, one of the sharpest increases in Europe during that decade. [25] Official Turkish data then show a prison population of 200,727 at the end of 2016, 291,546 in 2019, 341,294 in 2022, 383,663 in 2024 and 401,519 in 2025. [10]
The trajectory was interrupted—but not reversed—by pandemic-era execution-law changes. At the end of 2020, the prison population fell to 266,831, and the CPT recorded an even deeper temporary decline to around 192,000 at the time of its January 2021 visit after conditional-release, probation and pandemic-leave measures. [26] The subsequent rebound was striking: 337,760 prisoners by 3 June 2024, 392,456 by January 2025, 401,519 by year-end 2025 and 437,889 by November 2026. [27]
The capacity trend has been much slower. Official annual figures show capacity rising from 202,675 places in 2016 to 305,086 in 2025, while the prison population more than doubled over the same period. [10] The Turkish authorities argue that “operational capacity” is flexible, because extra beds can be installed and overnight accommodation remains available; the SPACE methodology records that qualification explicitly. [11]
The CPT, however, argues that the arithmetic itself demonstrates the limits of construction-led policy. During its February 2024 inspection, two L-type prisons designed for roughly 1,600 and 1,756 people were each holding almost 2,400, placing pressure on health care, activities and staff supervision. Its conclusion was unusually direct: without reducing admissions and time served, continually building prisons “will inevitably fail” to end overcrowding. [28]
Türkiye’s own strategy confirms that construction remains central. The authorities informed the CPT that dozens of new prisons were planned, while CTE says 52 institutions intended to add 45,289 places were envisaged over a five-year period. [29] These investments may improve physical conditions and replace antiquated buildings, but on the recent trajectory new capacity has been absorbed rapidly by population growth.
Pretrial detention requires more nuance than aggregate rhetoric sometimes allows. Türkiye’s share of remand prisoners is substantially below its level a decade ago: official year-end figures put remand prisoners at 36.2% of the prison population in 2016, falling to 12.4% in 2022 before rising to 15.9% in 2025 and about 16.5% in November 2026. [30] Yet the absolute number is now enormous—more than 72,000—and European institutions continue to criticise the frequency and legal quality of detention, particularly in politically sensitive cases, rather than merely its percentage share. The European Commission’s 2026 report says widespread and frequent use of pretrial detention persists. [31]
The European Court’s case law illustrates why. In Selahattin Demirtaş v. Türkiye (no. 4), the Court found that the relevant period of detention lacked reasonable suspicion and criticised the reasoning used to prolong detention for years. [32] In August 2026, the Grand Chamber in Kavala v. Türkiye (No. 2) found violations of Article 5 concerning arbitrary pretrial detention and post-conviction imprisonment, and held that the criminal proceedings had suffered such serious defects that they amounted to a flagrant denial of justice. [33]
These cases should not be generalized mechanically to every remand prisoner. They do, however, reinforce a systemic concern: overcrowding is partly generated upstream, by sentencing, detention and prosecutorial policy, rather than merely by prison administration.
Life inside: health, isolation and vulnerable prisoners
The CPT’s most recent published prison inspection, conducted in February 2024 and released in October 2026, offers the best independent institutional snapshot. Material conditions in the four high-security prisons it visited were “generally satisfactory,” but some prisoners at Iğdır S-type were sleeping on floor mattresses because beds were unavailable. More consequentially, the CPT found activity regimes in high-security institutions “highly unsatisfactory”: outside limited education and vocational programmes, prisoners received very little organised out-of-cell activity. [34]
The position of prisoners serving aggravated life sentences was especially severe. The CPT found many locked alone for at least 21 hours a day, sometimes for months or years, with solitary outdoor exercise as their only regular activity. It described this as small-group isolation or de facto solitary confinement and urged Türkiye to guarantee at least two hours of meaningful human contact daily. More fundamentally, it called the statutory aggravated-life regime itself “fundamentally flawed.” [28]
This is not simply a dispute over terminology. Türkiye’s rules provide aggravated-life prisoners with highly restricted visiting and telephone arrangements; the government response to the CPT explained that such prisoners may receive two visits per month and, subject to administrative approval, one ten-minute telephone call every fifteen days to a limited circle of relatives. The authorities also cited legal provisions permitting educational, cultural and sporting activities and described rehabilitation programmes for high-risk prisoners. [35]
A separate issue is irreducible life imprisonment. In its 2026 follow-up to Türkiye, the UN Committee against Torture noted government information indicating that about 4,000 people were serving aggravated life sentences and remained concerned that some categories still had no prospect of release. [36] The ECtHR’s August 2026 Kavala (No. 2) judgment independently found that the aggravated life sentence imposed there, without a realistic prospect of release or review, violated Article 3. [33]
Health care presents a similarly mixed picture. Turkish law guarantees examination and treatment, first through prison infirmaries and, where necessary, state or university hospitals. [37] The government has expanded prison health staffing, family-physician services and telemedicine, and told Parliament that prison infirmaries carried out roughly 4.5 million examinations and 1.817 million hospital referrals in 2025. [38]
Yet the CPT found insufficient health staffing in facilities it visited, shortcomings in psychiatric and psychological care, problems with medical confidentiality, and weaknesses in injury recording after alleged violence. [28] Civil-society estimates are necessarily incomplete but point in the same direction: the Human Rights Association (İHD) said it had identified at least 1,412 sick prisoners by April 2025, including 335 it categorised as seriously ill, while stressing that its figures covered only cases it could reach. [39] These numbers should be treated as a case inventory, not as a national prevalence estimate.
Deaths deserve equally careful handling. A ruling-party MP stated in Parliament that 1,001 prisoners died during 2025: 792 sentenced prisoners and 77 remand detainees in deaths classified as natural, plus 132 other deaths encompassing accidents and deaths during leave. He also reported 26 suicides between 1 January and 6 May 2026, 14 involving people who had been taking psychiatric medication. [13] These are official figures presented in parliamentary debate, but there is still no routinely published CTE mortality dataset with cause, age, institution, medical history and independent investigation outcome, limiting meaningful comparison.
Women make up a minority but a rapidly growing one. They were 4.7% of prisoners at the end of 2025, up from 4.2% in 2016. [12] During its visit to Antalya L-type Prison, the CPT found women’s accommodation affected by overcrowding and poor hygiene and concluded that mothers and small children were not being housed appropriately. It recommended a dedicated mother-and-baby unit and said mothers should never be placed in solitary confinement together with their children. [28] Turkish law permits children aged up to six to live with imprisoned mothers when no suitable outside caregiver is available, and the prison administration describes preschool, nutrition and psychosocial support arrangements for them. [40]
Children are another acute concern. As of 2 November 2026 there were 5,121 prisoners under 18; remarkably, 3,809—about three quarters—were remand detainees rather than sentenced children. [9] This is one of the clearest areas where alternatives to detention merit stronger use, given both the presumption of innocence and international juvenile-justice standards.
For foreign nationals, official year-end 2025 figures counted 15,759 prisoners, 3.9% of the prison population. [12] The CPT’s 2024 inspection of Maltepe L-type No. 3, which houses many foreign nationals, stressed the need for adequate staffing and training tailored to foreigners’ needs. [28]
For LGBTI+ prisoners, national data are especially poor. The CPT found that Turkish authorities had taken steps to provide transgender prisoners with safer accommodation, but also criticized practices that could produce months of de facto solitary confinement while gay prisoners awaited medical certification and urged adoption of clear protocols consistent with contemporary standards. [41] Civil-society “current statistics” still rely on an official figure of 200 LGBTI+ prisoners dating from 2018, which illustrates the data gap rather than the actual 2026 population. [42]
Older and disabled prisoners are likewise poorly captured in current public datasets. CTE counted 6,724 prisoners aged 65 or above in May 2026 and around the same order of magnitude thereafter. [43] Parliamentary material cited by CİSST listed 476 prisoners with specified disabilities in 2025, but those administrative categories are narrow and should not be read as a comprehensive disability census. [42]
Rehabilitation, labour and execution-law reform
One area where Türkiye’s prison system is more institutionally developed than some critical accounts suggest is education and work. CTE describes a continuum from literacy to open secondary education, university study, vocational training, libraries, sport and psychosocial programmes. [44] The İşyurtları system operates prison workshops and enterprises across more than 200 occupational fields, including agriculture, furniture, textiles, metalwork, food and construction-related production. [45]
Participation has expanded sharply. The İşyurtları 2025 annual report recorded 77,186 prisoners participating in work during that year, up from 58,193 in 2024. [14] By 30 October 2026, the Justice Ministry said roughly 82,000 prisoners had participated during 2026, with more than 32,000 participating at that moment. [15] Participants receive daily remuneration and insurance coverage for occupational accidents, occupational disease and maternity; the system is officially framed as vocational rehabilitation rather than ordinary employment. [46]
The most important policy question is therefore not whether rehabilitation programmes exist—they clearly do—but whether overcrowding and security classification restrict meaningful access. The CPT’s high-security findings suggest they often do: in institutions it visited, only limited numbers participated in education or vocational courses, while most prisoners had very little communal activity. [28] Rehabilitation capacity consequently appears unevenly distributed by prison type and security regime.
Execution-law reforms have repeatedly been used as a pressure valve. Pandemic-era legislation reduced time served before conditional release and expanded probation; a July 2023 measure facilitated early transfer to open prisons and probation, while tens of thousands placed on Covid leave did not return to closed institutions. [28]
The Tenth Judicial Package, enacted as Law No. 7550 in June 2025, changed conditional-release ratios, probation periods and special modes of sentence execution. [47] The Eleventh Judicial Package, enacted in December 2025 as Law No. 7571, extended earlier access to open prisons and probation to certain prisoners whose offences had been committed by 31 July 2023 but whose convictions became final later. The government presented the measure not as an amnesty but as a way to remedy unequal treatment caused by differing trial durations. [48]
Certain serious offences—including specified murders, sexual offences, terrorism and organised crime—were excluded. [49] By 7 January 2026, the Justice Ministry said 38,057 people had benefited. [50] The fact that the prison population nevertheless reached 437,889 ten months later is strong evidence that periodic release mechanisms alone are not counteracting the rate of new admissions and sentence accumulation. [8]
The CPT has identified another structural contributor: Turkish courts frequently impose consecutive sentences for multiple elements of what may constitute one criminal episode, producing aggregate terms substantially longer than in systems that apply greater concurrency. [28] That observation deserves more attention than the recurrent public debate over “amnesties,” because it points toward the sentencing architecture itself.
Political imprisonment, terrorism law and the judicial system
There is no universally accepted statistical category of “political prisoner” in Türkiye, and credible reporting should avoid pretending otherwise. People described by rights organisations as political prisoners are ordinarily recorded by the state according to criminal charges—terrorism, organisation membership, offences against the constitutional order, protest-related offences, disinformation or other Penal Code provisions. The analytical issue is therefore whether those laws and their application satisfy legality, evidence, necessity and proportionality standards.
Türkiye’s official position is that it faces genuine and continuing terrorism threats and that prosecution is directed at criminal acts and organisations rather than political opinion. The government designates the PKK and the Gülen movement, among others, as terrorist organisations and maintains that anti-terrorism enforcement is necessary for public security. Human-rights critics do not dispute the state’s right to prosecute violence; their central argument is that Turkish terrorism law is defined and applied too broadly to non-violent conduct. [51]
That criticism is no longer confined to NGOs. The European Commission’s November 2026 Türkiye report states that anti-terrorism legislation remains insufficiently aligned with European standards and continues to be applied disproportionately; it also reports continued investigations, prosecutions and convictions of journalists, lawyers, academics, human-rights defenders and other critical voices. [52]
Kurdish politicians and activists remain at the centre of this problem. Former HDP co-chairs Selahattin Demirtaş and Figen Yüksekdağ have been imprisoned since 2016. Human Rights Watch argues that numerous Kurdish politicians and activists have been prosecuted for non-violent political activities under terrorism provisions. [53] The ECtHR’s Demirtaş judgments have repeatedly found defects in the legal basis and purpose of detention; the 2025 judgment in Demirtaş (no. 4) again found a lack of reasonable suspicion during the relevant period. [32]
The context shifted politically in 2025–26 with renewed efforts to end the conflict with the PKK, Abdullah Öcalan’s February 2025 call for disarmament, the PKK’s subsequent announcement that it would dissolve and disarm, and the creation of a parliamentary process to consider related reforms. Human Rights Watch noted, however, that this process had not yet produced comprehensive reform of anti-terror legislation or a general review of cases involving imprisoned Kurdish politicians. [53]
The Osman Kavala litigation is even more consequential for the rule of law because it links detention directly to judicial independence. In August 2026 the ECtHR Grand Chamber found violations of Articles 3, 5 and 6 in Kavala v. Türkiye (No. 2) and concluded that serious defects in the proceedings—including issues affecting judicial independence and impartiality—had produced a flagrant denial of justice. [33] The European Commission reported that a Turkish court nevertheless rejected Kavala’s release request in October 2026. [54]
Journalists remain another measurable group, though figures vary according to definitions and dates. Amnesty International cited the Journalists’ Union of Türkiye as counting 14 journalists and media workers in prison in March 2026. [55] The European Commission, using a wider flow measure rather than a point-in-time prison count, reported that 74 journalists or media workers were arrested, imprisoned or remained in pretrial detention between September 2025 and September 2026. [56] These figures are not contradictory: one measures prison stock at a particular date; the other measures people affected over a twelve-month period.
Civil-society and legal professionals have also faced proceedings. The Commission’s 2026 report records prosecutions of human-rights defenders and pressure on civil-society organisations; Human Rights Watch documents cases involving lawyers representing defendants in politically sensitive proceedings. [57]
The institutional backdrop is critical. The European Commission concluded in November 2026 that Türkiye’s judiciary remained at an “early stage of preparation,” that backsliding had continued, and that executive influence over the judiciary undermined independence and the quality of judicial outcomes. It specifically called for reform of the Council of Judges and Prosecutors and greater compliance by lower courts with Constitutional Court and ECtHR judgments. [31]
This is where prison policy and constitutional governance converge. A prison system can be physically modern, offer vocational programmes and still produce rights violations if the process sending people into it is not sufficiently independent, evidence-based and reviewable.
Monitoring, accountability and what should change
Türkiye has a surprisingly dense formal monitoring architecture. Prison monitoring boards were established under Law No. 4681; TİHEK acts as the national preventive mechanism under the Optional Protocol to the Convention against Torture; the Ombudsman receives prisoner complaints; prosecutors exercise supervisory functions; and Türkiye is subject to CPT inspections, UN treaty-body review, ECtHR jurisdiction and Council of Europe supervision of judgments. [58] In 2023, prison monitoring boards reportedly carried out 2,054 visits to 381 penal institutions and produced 1,236 reports. [59]
The problem is less the absence of institutions than doubts over independence, transparency and follow-through. The UN Committee against Torture remained concerned in 2024 that complaints could still pass through prison administrations and that Türkiye lacked a sufficiently independent mechanism for investigating torture and ill-treatment. [60] The European Commission’s 2026 report similarly stated that Türkiye had not brought TİHEK’s national-preventive-mechanism role fully into line with OPCAT requirements. [61]
CPT transparency has improved but remains incomplete. Türkiye authorised publication in October 2026 of reports on the 2021 and 2024 prison visits and the 2022 İmralı visit, but the report on the CPT’s March 2025 visit—focused partly on detention following mass protests—had still not been published when the European Commission issued its November 2026 report. [62]
Civil-society monitoring fills important gaps but has methodological constraints. İHD’s prison reports are based on applications, correspondence, lawyer interviews and monitoring visits, and expressly warn that their cases do not constitute a comprehensive census of violations. Its Ankara branch, for example, recorded allegations affecting at least 577 prisoners during 2025, while the national report described recurring complaints involving medical access, isolation, searches, disciplinary procedures, communication restrictions and other conditions. [63] Such findings are best described as documented allegations and case patterns, not proof that every reported incident occurred exactly as alleged.
Timeline of the present system
2005–2015: Türkiye’s imprisonment rate rose from roughly 76 to 220 per 100,000, one of Europe’s sharpest increases. [64]
2016: Prison population passed 200,000 amid the broader post-coup criminal-justice expansion. [10]
2019: Population reached 291,546. [10]
2020–21: Pandemic-era release, probation and execution-law measures temporarily reduced the prison stock; the CPT observed around 192,000 prisoners during its January 2021 visit. [28]
July 2023: New execution provisions expanded early transfer to open prisons and probation for specified categories. [28]
June 2024: Prison population reached 337,760; the CPT warned that growth had resumed rapidly. [28]
July–August 2024: UN Committee against Torture reviewed Türkiye and raised concerns over torture investigations, detention conditions and aggravated life sentences. [65]
January 2025: SPACE I recorded 392,456 prisoners, a 458.1-per-100,000 imprisonment rate and 130.9 prisoners per 100 places. [11]
June 2025: Law No. 7550, the Tenth Judicial Package, revised conditional release, probation and special execution procedures. [66]
December 2025: Law No. 7571, the Eleventh Judicial Package, extended early transfer and probation rights to additional offenders whose crimes predated 31 July 2023 while excluding specified serious offences. [67]
August 2026: ECtHR Grand Chamber delivered Kavala v. Türkiye (No. 2), finding Articles 3, 5 and 6 violations. [33]
October 2026: CPT published long-awaited reports on its 2021, 2022 and 2024 Türkiye visits; its 2024 report criticised severe overcrowding, isolation and aspects of prison health care while also recording cooperation and generally satisfactory material conditions in several high-security facilities. [68]
2 November 2026: CTE recorded a new prison population of 437,889. [8]
Policy priorities
A credible reform agenda should begin with population reduction rather than capacity expansion alone. Türkiye should publish a prison-population strategy with measurable targets for lowering admissions, sentence length and remand use, alongside facility-level capacity and occupancy data. This follows the CPT’s repeated conclusion that construction cannot sustainably resolve overcrowding without changing sentencing and detention policy. [28]
Pretrial detention should be made demonstrably exceptional, particularly for children and non-violent defendants. Courts should be required to record individualized reasons explaining why bail, reporting requirements, electronic monitoring or other judicial-control measures are inadequate, with accelerated appellate review for prolonged detention. This would respond directly to ECtHR findings and Council of Europe standards on remand custody. [69]
The aggravated-life regime should be revised so that all prisoners have a realistic possibility of review and release based on rehabilitation and risk. De facto solitary confinement lasting months or years should end, with at least two hours of meaningful human contact daily as recommended by the CPT. [70]
Health data should become routinely public. CTE should publish annual facility-level statistics on deaths, suicide, self-harm, chronic disease, disability, hospital referrals, waiting times and causes of death, with independent investigation of deaths in custody. The absence of such disaggregated data currently makes both official reassurance and civil-society criticism harder to evaluate. [71]
Women, children, LGBTI+ people, foreign nationals, elderly prisoners and people with disabilities require separate policy frameworks and current statistics. In particular, detention of children awaiting trial should be sharply reduced; mother-and-child facilities should comply with the Bangkok Rules and child-development standards; and no prisoner should be placed in prolonged isolation merely because ordinary accommodation is regarded as unsafe for their sexual orientation or gender identity. [72]
Rehabilitation programmes should be protected from overcrowding. Work and vocational training are substantial assets in the Turkish system, but access should be audited by security classification, gender, disability and prison type, and remuneration and social-insurance arrangements should be reviewed against European prison standards. [73]
Finally, prison reform cannot be separated from judicial reform. Türkiye should implement binding ECtHR judgments, reform the structure and appointment process of the Council of Judges and Prosecutors to strengthen independence, narrow terrorism offences so that non-violent expression and association are not treated as evidence of organisational membership without compelling proof, and ensure that lower courts give effect to Constitutional Court judgments. These are not peripheral liberal reforms; they affect who enters prison, for how long and under what evidentiary threshold. [74]
The value of Prison Agenda is clearest in this context. Its reporting does not reveal a completely separate reality from the official record; rather, at its best it makes the official record legible, connects raw CTE statistics to European standards and draws attention to groups that disappear inside aggregate totals. Its quantitative claims on overcrowding and incarceration generally converge with independent Council of Europe evidence. Its rights-based interpretation, however, should remain distinguishable from verified facts and tested against primary sources, especially where individual allegations, causation or systemic intent are at issue. [75]
The broader evidence is difficult to dispute. Türkiye’s prisons in late 2026 contain more people than at any previous point in the modern statistical series; the country already had Europe’s highest comparable imprisonment rate before the additional growth of 2026; and both Turkish and international evidence shows that the system simultaneously possesses substantial modern infrastructure and rehabilitation programmes and suffers from chronic overcrowding, excessive isolation in parts of the high-security estate, gaps in medical and mortality transparency, and profound rule-of-law disputes over how some prisoners came to be there. [76]
For policymakers, the decisive question is no longer whether Türkiye can build enough prison beds. The past decade suggests that it cannot build its way out of its incarceration trajectory. The more consequential question is whether criminal-law, sentencing, detention and judicial institutions can be redesigned so that imprisonment becomes genuinely exceptional, proportionate and reviewable—while preserving public safety and the legitimate prosecution of serious crime. That is the point at which prison policy becomes a test of the rule of law itself. [5]
[1] [8] [76] # CİNSİYETE VE İNFAZ DURUMUNA GÖRE CEZA İNFAZ KURUMLARI MEVCUTLARI
[2] Türkiye/Turkey | World Prison Brief
https://www.prisonstudies.org/country/turkey?utm_source=chatgpt.com
[3] Adrese Dayalı Nüfus Kayıt Sistemi Sonuçları – 2025 – Veri Portalı – TÜİK
[4] [11] [16] Prison stock on 31 January 2025 – Council of Europe Annual Penal Statistics
[5] [6] [7] [27] [28] [34] [41] [70] rm.coe.int
[9] [72] # 2.11.2026 TARİHİ İTİBARIYLA CEZA İNFAZ KURUMLARINDA BULUNAN TUTUKLU VE HÜKÜMLÜLERİN ÖĞRENİM DURUMLARINA GÖRE DAĞILIMLARI
[10] [12] [30] Ceza infaz kurumlarının 31 Aralık tarihi itibarıyla sayısı, kapasitesi, personel durumu ve nüfusu, 2016-2025
[13] [71] TÜRKİYE BÜYÜK MİLLET MECLİSİ
https://tbmm.gov.tr/Milletvekili/UyeGenelKurulKonusmalariDetay?eid=144865&utm_source=chatgpt.com
[14] [73] Cezaevlerinin terbiye, ıslah ve iş esaslarına göre düzeltilmesi yolundaki hayırlı
[15] Türkiye Cumhuriyeti Adalet Bakanlığı
[17] [75] About – Prison Agenda
https://prisonagenda.substack.com/about?utm_source=chatgpt.com
https://prisonagenda.substack.com/p/prison-agenda-1?utm_source=chatgpt.com
[20] [23] Türkiye in the CoE’s Annual Criminal Justice Statistics
https://prisonagenda.substack.com/p/turkiye-coes-annual-criminal-2025?utm_source=chatgpt.com
[21] Prison Agenda #6 – by Ahmet Kavruk – Prison Agenda
https://prisonagenda.substack.com/p/prison-agenda-6?utm_source=chatgpt.com
[22] Prison Agenda | Substack
https://prisonagenda.substack.com/t/article?utm_source=chatgpt.com
[24] The people behind the numbers: Bringing the reality of Turkish prisons to light – US News Hour
[25] [64] Ten-year trends in European prisons – Portal
https://www.coe.int/en/web/portal/-/ten-year-trends-in-european-prisons?utm_source=chatgpt.com
[26] Prison Statistics – 2020 – Data Portal – TURKSTAT
https://veriportali.tuik.gov.tr/en/press/37202/metadata?utm_source=chatgpt.com
[31] [74] inclusive and evidence-based methods.
[32] [69] Selahattin Demirtaş v. Türkiye (no. 4)
https://hudoc.echr.coe.int/fre?i=002-14489&utm_source=chatgpt.com
[33] Kavala v. Türkiye (No. 2) [GC]
https://hudoc.echr.coe.int/rus?i=002-14651&utm_source=chatgpt.com
[36] Letterhead_OHCHR
[37] UYAP Mevzuat ve İçtihat Programı
https://mevzuat.adalet.gov.tr/mevzuat/102944?query=5275&utm_source=chatgpt.com
[39] | | | | 4
[40] SSSorularDetay
https://cte.adalet.gov.tr/Home/SSSorularDetay/11?utm_source=chatgpt.com
[42] Kasım 2026 Hapishane İstatistikleri – CİSST
https://cisst.org.tr/kasim-2026-hapishane-istatistikleri/?utm_source=chatgpt.com
[43] Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla Yaş Gruplarına Göre Dağılım | Hükümlü | | | Tutuklu | | | GENEL TOPLAM | | — | — | — | — | — | — | — | — | — | — Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla None | Erkek | Kadın | TOPLAM | Erkek | Kadın | TOPLAM | Erkek | Kadın | TOPLAM Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla 12-18 Yaş Arası | 1.215 | 78 | 1.293 | 3.250 | 137 | 3.387 | 4.465 | 215 | 4.680 Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla 18-40 Yaş Arası | 219.182 | 10.368 | 229.550 | 42.408 | 2.877 | 45.285 | 261.590 | 13.245 | 274.835 Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla 40-65 Yaş Arası | 114.624 | 5.687 | 120.311 | 13.201 | 1.047 | 14.248 | 127.825 | 6.734 | 134.559 Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla 65 Yaş ve Üzeri | 5.900 | 229 | 6.129 | 568 | 27 | 595 | 6.468 | 256 | 6.724 Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla GENEL TOPLAM | 340.921 | 16.362 | 357.283 | 59.427 | 4.088 | 63.515 | 400.348 | 20.450 | 420.798 Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin Yaş Gruplarına Göre Dağılımları 04.05.2026 Tarihi İtibarıyla Bu tablo, 19/04/2008 tarihli ve 26852 sayılı Resmî Gazete’de yayımlanan 2008/13472 sayılı Bakanlar Kurulu Kararı ile yürürlüğe giren Resmî İstatistik Programı kapsamında olmayıp, hizmete özel olarak hazırlanmıştır. | | | | | | | | |Ceza İnfaz Kurumlarında Bulunan Tutuklu ve Hükümlülerin
[44] Eğitim ve Öğretim Faaliyetleri
https://cte.adalet.gov.tr/Home/SayfaDetay/egitim-servisi-faaliyetleri?utm_source=chatgpt.com
[45] Türkiye Cumhuriyeti Adalet Bakanlığı
[46] 2025 yılı performans özgür 2025_27_01_2025.indd
[47] [66] Mevzuat Genel Müdürlüğü
https://mgm.adalet.gov.tr/Home/?utm_source=chatgpt.com
[48] [49] [67] Türkiye Cumhuriyeti Adalet Bakanlığı
[50] ADALET BAKANI YILMAZ TUNÇ, MECLİS’TE GAZETECİLERİN SORULARINI YANITLADI
[51] Advancing Human Rights, Justice and Democracy for Kurds and All Other Communities in Türkiye | Human Rights Watch
[52] [61] IMMC.SWD%282026%29696%20final.ENG.xhtml.8_EN_autre_document_travail_service_part1_v7.docx
[53] World Report 2026: Türkiye | Human Rights Watch
https://www.hrw.org/world-report/2026/country-chapters/turkiye?utm_source=chatgpt.com
[55] Türkiye: End the crackdown on journalists – Amnesty International
https://www.amnesty.org/en/documents/eur44/0957/2026/en/?utm_source=chatgpt.com
[56] Between September 2025 and September 2026, according to CSOs working in media freedom 74 journalists and media workers were arrested, imprisoned, or remained in pre-trial detention. (51 in the same period in 2024-2025).
[58] TİHEK – ULUSAL ÖNLEME MEKANİZMASI
[59] İŞKENCE VE KÖTÜ MUAMELEYE KARŞI ULUSAL ÖNLEME MEKANİZMASI
https://tihek.gov.tr/public/editor/uploads/1BdzBwlN.pdf?utm_source=chatgpt.com
[60] Docstore
[62] The CPT and Türkiye – CPT
https://www.coe.int/en/web/cpt/turkiye
[63] İHD Ankara Şubesi Hapishaneler Komisyonu 2025 Yılı Hapishanelerde Hak İhlalleri Raporu | İnsan Hakları Derneği – İHD
[65] Docstore
[68] Council of Europe anti-torture Committee (CPT) publishes two reports on Türkiye focusing on police and prisons – CPT
Discover more from Erkan's Field Diary
Subscribe to get the latest posts sent to your email.
