The UN Universal Periodic Review of China Across Four Cycles (9 February 2009 to 23 January 2024): 428 Recommendations in the Fourth Cycle, 290 Accepted and 98 Rejected
This page records all four Universal Periodic Review cycles of China, from 252 recommendations in 2013 and 346 in 2018 to 428 on 23 January 2024, of which China accepted 290 and rejected 98, as its acceptance rate fell from about 82 percent to about 68 percent.
1. Introduction: a peer review no state can avoid
The Universal Periodic Review is the most universal mechanism of the United Nations Human Rights Council, and its defining feature is that it admits no exceptions. Large or small, whatever its record, every member state of the United Nations must periodically take the seat in Geneva, listen to recommendations delivered one by one by other states, and declare, on the record, what it accepts and what it refuses. China, a permanent member of the Security Council and a repeatedly elected member of the Council itself, is not exempt. This page records the four cycles China has undergone between 2009 and 2024: the first on 9 February 2009, the second on 22 October 2013, the third on 6 November 2018 and the fourth on 23 January 2024. Across fifteen years the number of recommendations rose from 252 in the second cycle and 346 in the third to 428 in the fourth, while China's acceptance rate fell back sharply in the fourth cycle and its rejections reached an all-time high. That pair of curves, one rising and one falling, is the key to this page.
The legal character of the mechanism must be stated plainly at the outset. Recommendations under the Universal Periodic Review are not legally binding. A state under review may accept them, partially accept them, note them or reject them, and rejection breaches no treaty obligation. It does not follow that the review is a formality. Each cycle produces formal documents recording what was recommended and how the state replied. Accepted recommendations become public commitments to the international community, and at the next cycle states ask, item by item, what has been done. Rejected and merely noted recommendations also remain in the file, marking the precise list of undertakings a government declined to give. Read together, the four cycles form a history of the human rights agenda as written jointly by the world's governments and signed for, or returned unsigned, by Beijing. This page assembles the dates, totals, acceptance and rejection patterns, substantive focal points and follow-up for all four cycles from official United Nations document symbols, and flags honestly every point where sources diverge rather than filling gaps with unverified figures.
The substantive centre of gravity also moved across the cycles. Early reviews concentrated on institutional questions such as timetables for ratifying international covenants, the death penalty and re-education through labour. By the third cycle Xinjiang had become a focal point for the first time, with multiple states demanding the closure of facilities described as re-education camps. By the fourth cycle, recommendations concerning Hong Kong, Tibet and Xinjiang together numbered 57, of which China accepted only 19, roughly one third. The questions also grew more specific. Six states, France, Luxembourg, the United Kingdom, the United States, Sweden and Australia, called for the abolition of residential surveillance at a designated location. Recommendations invoking the Xinjiang assessment of the Office of the High Commissioner for Human Rights were rejected on the ground that the report was completely illegal and invalid. A single review might be dismissed as diplomatic ritual. A curve drawn through four cycles cannot: the world's states asked China more questions, and more precise questions, while the share Beijing was willing to commit to shrank markedly in the fourth round.
2. The mechanism and its procedure
The Human Rights Council was created in 2006, and the Universal Periodic Review was established at its foundation as a core mechanism covering all 193 member states of the United Nations, each reviewed roughly every four and a half years. Reviews take place in a working group, where delegations speak in turn and the state under review responds on the spot. A working group report then lists every recommendation and the state's preliminary position. The state later files an addendum giving its final reply to each item, and the Council formally adopts the outcome in plenary. Reply categories include accepted, partially accepted, noted and rejected. Acceptance signals a commitment to implement; partial acceptance commits the state to only part of a recommendation; noting occupies ambiguous ground between acceptance and refusal; rejection is an explicit refusal. States have not used these categories uniformly across cycles. In its earlier cycles China worked mainly with the vocabulary of acceptance and non-acceptance, while in the fourth cycle it used rejection on a large scale for the first time. The change of vocabulary is itself evidence of a change of posture.
Three bodies of material ground each review: the national report of the state under review, a compilation of United Nations information prepared by the Office of the High Commissioner for Human Rights, and a summary of submissions by stakeholders including non-governmental organisations. A review is therefore not a mere exchange of diplomatic courtesies. Behind each delegation's statement sit treaty body conclusions, special procedures reports and civil society documentation. In China's case, the successive conclusions of the Committee on the Elimination of Racial Discrimination, the Committee Against Torture and the Committee on Economic, Social and Cultural Rights, together with the OHCHR Xinjiang assessment released on 31 August 2022, supplied the factual basis for many of the questions put in later cycles. This also explains the vehemence with which recommendations citing the Xinjiang assessment were refused in the fourth cycle: Beijing was rejecting not a single recommendation but an entire chain of questioning anchored in the United Nations' own assessment.
A further procedural element is mid-term follow-up. States are encouraged, though not required, to submit a voluntary mid-term implementation report between cycles, accounting for progress on accepted recommendations. Whether such a report is filed is itself an indicator of attitude. For the fourth cycle, as of this page's latest check, China has filed no official mid-term implementation report, and no itemised public accounting of the accepted recommendations has appeared. On 29 September 2026 the British human rights organisation The Rights Practice published a mid-term assessment of the fourth cycle. Whenever this page cites that milestone it labels it for what it is: an assessment by a non-governmental organisation, not a United Nations conclusion, and the two carry different evidentiary weight. Beyond that publication, this page found no new official review milestone after the adoption of the fourth-cycle outcome on 4 July 2024, and the account of the latest status below is bounded accordingly and stated as such.
3. The first and second cycles: from 2009 to 2013
The first-cycle review took place on 9 February 2009, at the fourth session of the Working Group on the Universal Periodic Review, and the Human Rights Council adopted the outcome at its eleventh session in June 2009 under the symbol A/HRC/11/25. It was the first time China had faced peer scrutiny from the entire membership under this mechanism. Sources diverge on the total number of recommendations made in that first cycle, and this page deliberately adopts no unverified total, recording the outcome qualitatively instead: China accepted most of the recommendations addressed to it, while rejecting those concerning the death penalty and the setting of a timetable for ratifying the International Covenant on Civil and Political Rights, among others. That pattern, accepting the majority while refusing the core institutional questions, recurred in all three later cycles and was set in the very first round. Recommendations on economic development, social security, poverty reduction and concrete welfare measures were generally welcomed; anything touching ratification of the political rights covenant, the survival of the death penalty, or the core arrangements of the judicial and detention system was turned away.
The second-cycle review took place on 22 October 2013, and its figures rest on formal documents. China received 252 recommendations, accepted 204 and did not accept 48. The outcome documents are A/HRC/25/5 and its Addendum 1, and the Human Rights Council adopted the outcome on 20 March 2014. The recommendations not accepted clustered in familiar places: demands for a definite timetable for ratifying the International Covenant on Civil and Political Rights, which China signed in 1998 but has never ratified and which would be raised again in every subsequent cycle; abolition of the death penalty and related transparency demands; and the system of re-education through labour. One adjacent fact deserves recording with care: re-education through labour was in fact abolished at the end of 2013, shortly after the second-cycle review. The abolition, however, does not retroactively convert the review replies into acceptances, and this page keeps the two planes strictly separate, the answers given in Geneva on one side and later domestic policy change on the other.
Set side by side, the first two cycles reveal China's early posture within the mechanism. A high acceptance rate was its principal instrument for presenting itself as cooperative, and 204 acceptances against 48 refusals did keep the second-cycle acceptance share above four fifths. Yet the 48 refusals were never randomly distributed. They landed precisely on the sensitive band of political rights and judicial detention. This selective acceptance means that the acceptance rate alone cannot carry the interpretive weight often placed on it: a state can accept a large volume of social and welfare recommendations while saying no to every proposal that would constrain power. The documents of the first two cycles must therefore be read together with the content of the recommendations, not as bare percentages. This page applies the same reading to the later cycles, and in particular to the fourth, where both the falling rate and the specific targets of refusal require attention.
4. The third cycle: the 2018 review and the arrival of Xinjiang
The third-cycle review took place on 6 November 2018, with some 150 states registered to speak. China received 346 recommendations, accepted 284 and did not accept 62; its replies are recorded in A/HRC/40/6/Add.1, dated 15 February 2019. At roughly 82 percent, this was the highest acceptance rate of the four cycles. The historical significance of the round, however, lies not in the percentage but in the changed composition of the agenda: Xinjiang became one of the focal points of a review of China for the first time. Since 2017, reporting on the mass establishment of so-called Vocational Education and Training Centres in Xinjiang, and on the transfer into them of large numbers of Uyghurs and other Muslim minorities, had become widely known internationally. Multiple states used their statements to demand directly that China close the facilities described as re-education camps, end arbitrary detention and allow unimpeded access to United Nations personnel. Those recommendations were mostly rejected or merely noted and did not enter the accepted list. The exchanges over Xinjiang in that room foreshadowed the basic configuration of the Xinjiang dispute inside the United Nations system for years to come.
Other focal points of the third cycle continued the established lines. The timetable for ratifying the International Covenant on Civil and Political Rights was raised again by many states and again not accepted. The death penalty, freedom of expression and association, the treatment of lawyers and rights defenders, and questions concerning Hong Kong and Tibet all featured. In its statements China emphasised its human rights achievements, foregrounding poverty reduction, development and social security, and restated its standing position against the politicisation of human rights and interference in internal affairs. Procedurally, the third cycle supplied the largest number of acceptances, 284, and it remains the figure Beijing most often cites when describing its own constructive engagement with the review. Yet the same cycle changed the character of the exercise, because the questions were no longer confined to the institutional plane of covenants and statutes. They pointed at a specific, operating system of mass detention and at its victims. Everything that followed, the High Commissioner's assessment, the Council's debate vote, and the concentrated questioning of the fourth cycle, can be traced back to that first massed, face-to-face interrogation in November 2018.
The third cycle also resonated with parallel United Nations procedures. In 2018 the Committee on the Elimination of Racial Discrimination, reviewing China's treaty report, had already expressed concern at reports of mass detention in Xinjiang. The statements in the Universal Periodic Review and the treaty body's conclusions corroborated one another, showing that Xinjiang entered the United Nations agenda through several independent channels at once rather than as any single state's fabrication. China's standing reply was to describe the centres as lawful vocational education, training and de-radicalisation measures, and to host visits by selected delegations. This page records both presentations, but the procedural fact is unaffected by either: among the 346 recommendations received in the third cycle, not one directly addressing the Xinjiang detention facilities was accepted by China. That single line of record is the necessary starting point for understanding the sharper confrontation of the fourth cycle.
5. The fourth cycle: the 2024 review and the turning point in acceptance
The fourth-cycle review took place on 23 January 2024, at the forty-fifth session of the Working Group, with about 160 states registered to speak, the largest number across the four cycles. China received 428 recommendations, and the working group report carries the symbol A/HRC/56/6. China's itemised replies appear in the addendum A/HRC/56/6/Add.1, dated 31 May 2024: 290 recommendations accepted, 8 partially accepted, 32 noted and 98 rejected. The Human Rights Council adopted the fourth-cycle outcome at its fifty-sixth session on 4 July 2024, and Ambassador Chen Xu, head of the Chinese delegation, addressed the adoption. The raw count of acceptances, 290, was the highest of any cycle. As a proportion, however, acceptance stood at roughly 68 percent, a clear fall from roughly 82 percent in the third cycle. Still more significant was the vocabulary: 98 rejections marked the first large-scale use of the rejected category in China's review history, replacing the softer non-acceptance formula of earlier rounds. Numbers and wording turned at the same time, and that joint turn is the central fact of the fourth cycle.
The substantive distribution is more revealing than the headline rate. According to a joint analysis published on 31 July 2024 by five human rights organisations including the International Federation for Human Rights, of the 57 recommendations concerning Hong Kong, Tibet and Xinjiang, China accepted only 19, about 33 percent, barely a third. The decline in overall acceptance was therefore not evenly spread. It was heavily concentrated in the three territorial and ethnic questions under the closest international scrutiny, while recommendations on economic development and social rights continued to be accepted in large numbers. The five organisations criticised China for offering denials at odds with the evidence during the review. This page records that criticism with its source clearly labelled as a joint analysis by non-governmental human rights organisations, not as a conclusion of any United Nations body. China's public position is recorded alongside it: Beijing states that it accepted the vast majority of constructive recommendations and opposed only those it regards as politicised or based on false information. Both accounts stand in the file, and readers can test them against the itemised replies themselves.
Two specific features of the fourth cycle must be written out in full. First, six states, France, Luxembourg, the United Kingdom, the United States, Sweden and Australia, recommended the abolition of residential surveillance at a designated location, a coercive measure under which suspects may be held incommunicado at a designated place, long criticised by United Nations experts and human rights organisations as a form of de facto secret detention. That six governments jointly wrote this single institution into their recommendations shows it had become a shared target of concern. Second, several recommendations invoked the OHCHR Xinjiang assessment and urged China to implement it, among them the items at paragraphs 22.28 to 22.30 of A/HRC/56/6. China rejected them on the stated ground that the report was completely illegal and invalid, adding that the Human Rights Council had already voted down the related draft decision. The reference is to the vote of 6 October 2022, when the Council rejected by 17 votes to 19 a draft decision merely to hold a debate on Xinjiang, a vote recorded on its own page in this section of the archive. The logic of that reply, using one procedural vote to annul the authority of an assessment issued by the United Nations' own human rights office, is itself preserved here for the record.
6. Reactions and implementation: the position as of November 2026
Reactions around the adoption of the outcome followed established lines. States supportive of China praised its human rights achievements and poverty reduction in their statements and at the adoption, stressing respect for each country's chosen path of human rights development. States that had put sharp recommendations restated, in their explanations, their concerns over Xinjiang, Tibet and Hong Kong and over individual detained defenders. On 31 July 2024 five human rights organisations, the International Federation for Human Rights, Human Rights in China (HRIC), the International Campaign for Tibet, the Taiwan Association for Human Rights and the Database Center for North Korean Human Rights, published their joint analysis. Beyond the count of 57 recommendations and 19 acceptances, they criticised China for denials contradicted by evidence throughout the review and urged states to keep tracking the subject matter of the rejected recommendations in later procedures. This page cites that analysis for its statistical scope and its stated criticisms, with its character as non-governmental analysis made explicit at every mention. Its figures of 57 and 19 can be verified item by item against A/HRC/56/6 and its addendum.
The implementation record, by contrast, is thin. The Universal Periodic Review has no enforcement machinery. Delivery on accepted recommendations depends on voluntary reporting by the state and on retrospective questioning at the next cycle. For the fourth cycle, China has filed no official mid-term implementation report, and no itemised public account of progress on the 290 accepted recommendations has appeared in open sources. Beijing speaks in general terms of new progress in its human rights cause, but such formulations cannot be matched against particular recommendation items for verification. On 29 September 2026 The Rights Practice, a British human rights organisation, published a mid-term assessment of the fourth cycle examining implementation of the accepted recommendations. This page stresses once more that it is a non-governmental assessment and not a United Nations conclusion. Its value lies in supplying a checkable civil-society counter-text in the absence of an official report. Until the fifth cycle begins, these fourth-cycle documents, together with that assessment, constitute the whole public basis on which the international community can test whether China's undertakings were honoured.
As of this page's latest check in November 2026, no new cycle of the review of China has begun, and the next review awaits scheduling by the United Nations. Nor has the Human Rights Council created any dedicated follow-up mechanism for the rejected fourth-cycle recommendations. What the four cycles leave behind is therefore a static but complete archive: 428 fourth-cycle recommendations on the record, four reply categories recorded item by item, and the subjects of 98 rejections identified one by one. The purpose of this page is to fix the outline, the symbols and the correct way of reading that archive, so that before the next cycle opens anyone can establish quickly what the Chinese government once promised, what it refused, and which questions the world's governments have formally put to it more than once. An archive generates no obligations by itself, but it makes denial harder: a question asked across four cycles can no longer be described as a question never raised.
7. Documents and procedure index
The core documents are indexed by cycle as follows. Mechanism: the Universal Periodic Review was established with the Human Rights Council in 2006, covers all 193 member states, and runs on a cycle of roughly four and a half years. First cycle: review on 9 February 2009 at the fourth session of the Working Group; outcome adopted by the Council at its eleventh session in June 2009, symbol A/HRC/11/25; sources diverge on the first-cycle recommendation total, and no specific total is adopted on this page. Second cycle: review on 22 October 2013; 252 recommendations received, 204 accepted and 48 not accepted; documents A/HRC/25/5 and Addendum 1; outcome adopted on 20 March 2014. Third cycle: review on 6 November 2018 with about 150 states speaking; 346 recommendations received, 284 accepted and 62 not accepted; replies in A/HRC/40/6/Add.1 dated 15 February 2019. Fourth cycle: review on 23 January 2024 at the forty-fifth session of the Working Group with about 160 states registered to speak; 428 recommendations received; report A/HRC/56/6; addendum A/HRC/56/6/Add.1 dated 31 May 2024 recording 290 accepted, 8 partially accepted, 32 noted and 98 rejected; outcome adopted by the Council at its fifty-sixth session on 4 July 2024.
Supporting and follow-up materials include the joint analysis of five human rights organisations dated 31 July 2024, counting 19 acceptances among 57 recommendations on Hong Kong, Tibet and Xinjiang; the recommendations by France, Luxembourg, the United Kingdom, the United States, Sweden and Australia calling for abolition of residential surveillance at a designated location; and the recommendations invoking the OHCHR Xinjiang assessment, including paragraphs 22.28 to 22.30 of A/HRC/56/6, which China rejected on the ground that the report was completely illegal and invalid. On follow-up, China has filed no official fourth-cycle mid-term implementation report, and there is no public record of an itemised implementation list; the mid-term assessment published by The Rights Practice on 29 September 2026 is a non-governmental document. Every symbol and date in this index can be checked in the United Nations document system. Where this page could not verify a formal symbol or a settled figure, as with the first-cycle total, the gap is declared rather than filled by conjecture.
8. Verification note
The recommendation totals and the acceptance and non-acceptance figures for the second, third and fourth cycles are recorded from A/HRC/25/5 and Addendum 1, A/HRC/40/6/Add.1, and A/HRC/56/6 with A/HRC/56/6/Add.1 respectively. The acceptance rates of about 68 percent for the fourth cycle and about 82 percent for the third are approximations calculated from those figures and are worded as approximations throughout. Because sources diverge on the total number of first-cycle recommendations, this page, as instructed by its verified fact base, describes the first cycle qualitatively and gives no total; that is a deliberate, declared omission, not an oversight. The figures of 57 recommendations and 19 acceptances come from the joint analysis of five human rights organisations dated 31 July 2024 and are labelled as non-governmental analysis wherever they appear, never presented as United Nations statistics. The mid-term assessment by The Rights Practice dated 29 September 2026 is likewise labelled a non-governmental assessment. The statement that China has filed no official fourth-cycle mid-term report is framed as the absence of any public record as of this page's check, a falsifiable search result rather than an absolute assertion. No quotation on this page exceeds fifty words, and every figure and document symbol follows the verified fact base without alteration or extrapolation.
Our Take
The Universal Periodic Review is often mocked as a tea party of mutual congratulation, and China's four cycles suggest a more precise reading: the congratulation is real, but the itemised reply documents do not lie. In a single addendum, 290 acceptances stand printed beside 98 rejections. The first figure proves that Beijing does not refuse everything. The second traces with precision what it does refuse: only 19 of 57 recommendations on Hong Kong, Tibet and Xinjiang accepted, residential surveillance at a designated location retained, and the High Commissioner's Xinjiang assessment dismissed as completely illegal and invalid. The acceptance rate fell from 82 percent to 68 percent, and what fell away was not welfare policy but the politically sensitive band. This pattern of selective acceptance has run for fifteen years, and its information content lies in its stability: what a government will promise says far less about where it draws the boundary of power than what it persistently will not promise.
The fourth cycle also introduced, at scale, the blunt word rejected. Earlier replies had the buffers of non-acceptance and noting. The fourth cycle instead rejected 98 items outright, including recommendations grounded in an assessment by the United Nations' own human rights office. To annul that assessment's authority by citing a single 17-to-19 procedural vote is legally untenable, but as posture it is eloquent: Beijing's engagement with the international human rights machinery is shifting from selective cooperation toward selective confrontation. The reviews will continue, cycle after cycle, and the recommendations will keep being made. The point of the archive is that when the fifth cycle opens, no one will be able to pretend that the 98 rejections of the fourth never happened, or to present a question asked four times as a novel accusation.
9. Supplementary record: how the reply categories evolved
This supplementary record addresses a technical detail easily overlooked, the evolution of the reply categories, because that evolution is among the most honest curves in the four-cycle archive. In the first and second cycles China's replies were essentially binary, acceptance or non-acceptance, with noting used only occasionally and rejection not a principal category. The third cycle broadly preserved that binary structure, 284 acceptances against 62 non-acceptances, formally still a reply dominated by cooperation. In the fourth cycle the categories expanded into four working tiers: accepted, partially accepted with 8 items, noted with 32, and rejected with 98. Partial acceptance allows Beijing to take only the harmless portion of a composite recommendation. The wider use of noting supplies a buffer that neither promises nor openly refuses. The mass deployment of rejection isolates the most sensitive batch of recommendations in a category of their own. That all three instruments of buffering and isolation appear together indicates a reply worked over with legal and diplomatic care, not an improvised response.
The evolution carries a methodological instruction for later readers: do not merely count acceptances, but trace the reply trajectory of a single issue across all four cycles. Recommendations for a timetable to ratify the International Covenant on Civil and Political Rights recur in every cycle and were accepted in none, forming a complete trajectory of refusal; death penalty recommendations trace a similar line. The Xinjiang trajectory is shorter and steeper: the issue entered in the third cycle and was immediately rejected or noted, and in the fourth the recommendations, together with those citing the High Commissioner's assessment, were refused in the harshest available wording. Conversely, recommendations on poverty reduction, social security and specific rights of women and children were mostly accepted in all four cycles. Drawn item by item, these trajectories yield not a snapshot of one year but a map of where power limits itself. This page concentrates the document symbols of all four cycles in one index precisely so that any reader can follow that map and verify every line of it independently.
Two procedural cautions are appended for future checking. First, the numbers of states registered to speak and the numbers actually taking the floor may differ slightly in individual cycles; this page's figures of about 150 states for the third cycle and about 160 for the fourth follow the registered-speaker convention and are worded as approximations. Second, the finding that China has filed no official fourth-cycle mid-term report rests on the absence of any public record as of November 2026; should such a report be filed later, this page will incorporate it by substantive update with its date noted. The archive is preserved and never deleted.