Has anyone found errors in Gaza’s fatality lists?

Yes. On 29 July 2024 the Ministry of Health in Gaza, which operates under the Hamas-run authority and publishes its own count of the dead without independent verification, stated that a committee auditing dates of death had found decedents registered several months after they died, some of whom had been under 18 at death. The ministry’s own published figures also carry checkable slips: its split of the count for 30 June 2024 sums to 37,900, the total it reported for 1 July. The one peer-reviewed statistical test of the lists, a capture-recapture study in The Lancet (Jamaluddine and colleagues, 2025), estimated 64,260 traumatic-injury deaths to 30 June 2024 against the ministry’s own 37,877, a gap that runs toward undercounting rather than inflation.

Key Findings

• The Gaza Ministry of Health, a party to the war reporting its own count, stated on 29 July 2024 that an audit committee found decedents registered months after death, and on 24 July 2024 counted 9,715 deaths whose data was not yet complete as of 30 June 2024 (Ministry of Health, Gaza, 2024 to 2026).
• The ministry’s own split for 30 June 2024, 28,185 complete records plus 9,715 incomplete, sums to 37,900, the figure in its 1 July 2024 daily report, while its 30 June report gives 37,877; not independently verified (Ministry of Health, Gaza, 2024 to 2026).
• 5,817 of the 28,185 records in the ministry’s list to 30 June 2024, or 20.6%, carry its own label for notification by relatives rather than hospital records, against 2,478 of 21,323 (11.6%) in the list to 29 March 2024; a party’s own provenance field (Ministry of Health, Gaza, 2024 to 2026).
• A peer-reviewed capture-recapture study estimated 64,260 traumatic-injury deaths (95% CI 55,298 to 78,525) from 7 October 2023 to 30 June 2024, so 41% of the estimate was absent from the ministry’s own reported total of 37,877 (Jamaluddine and colleagues, 2025).
• For the 2014 war, the Gaza ministry’s count exceeded the Israeli Foreign Ministry’s by 8.0% on the Gaza ministry’s count as base, each a party’s own figure, as reported in a Lancet correspondence letter that is not peer-reviewed (Huynh, Chin and Spiegel, 2024).

Has the Gaza ministry acknowledged errors in its own lists?

Yes: the Ministry of Health in Gaza stated on 29 July 2024 (message 5663) that dates of death had been registered on a family’s report or on the day a body reached hospital, and that a committee formed to audit the actual dates found a number of decedents who had been under 18 at death but were registered several months later. The statement is a party’s own account, not independently verified.

The Foundation for Defense of Democracies, a Washington policy institute, published an analysis on 9 April 2024 under the headline “Hamas-run Gaza Health Ministry admits to flaws in casualty data,” and that framing, of the issuing body conceding defects in its own record, circulates widely. The ministry’s own statements bear out the concession in a specific form. The 29 July correction concerns when deaths were dated, not whether the listed people died. On the ministry’s own description, a death dated at registration rather than at the moment of death records the decedent as older than they were, so the committee’s correction runs toward the child count.

The ministry also publishes a standing count of what its named list does not yet cover. Its post of 24 July 2024 (message 5651) gave 9,715 deaths whose data was not yet complete as of 30 June 2024, beside 28,185 records with complete data, and defined the two: the general count is everyone who reached the hospitals, and the complete-data count is everyone whose data the ministry has finished. Both are the ministry’s own reported totals.

That distinction governs every later reading. A named list and a cumulative total are built differently, a point set out document by document in what the Gaza health ministry publishes about its own count. The ministry adds records to the cumulative total by administrative approval months after the deaths, so a later list can be longer than an earlier one without new deaths in between, and a total on a given date is never the same object as a list on that date.

What the ministry’s own statements establish, then, is a self-reported correction to dates of death, a self-reported backlog of incomplete records, and a published definition separating the two counts. None of it is a count of erroneous entries: the ministry’s posts cited here give no number of records corrected by the 29 July audit.

Who says names have vanished from the lists, and since when?

The Jewish Chronicle, a London weekly, and HonestReporting, a media-monitoring organization whose stated mission is countering what it describes as bias against Israel, reported in 2025, the latter through The Jerusalem Post, that about 3,400 identified deaths listed by Gaza’s Ministry of Health in August and October 2024 were absent from its March 2025 update. Both accounts put more than 1,000 children among the absent records.

The lists do change between rounds, and the ministry’s own channel records part of that movement. On 23 March 2025, in the same daily report that carried a cumulative 50,021 deaths, the ministry added 233 records it described as completed and approved by its judicial committee following family notifications and the missing-persons file (message 6377). It added 280 more on 30 August 2025 (message 6900) and 160 during August 2026 (message 7600), there attributing approval to a “martyrs’ approval committee.” These are the ministry’s own reported additions, not independently verified, and they enter the total by administrative approval rather than as new deaths.

An absence between 2 rounds is a count of names, not of reasons. A record missing from a later list can reflect a removal, a merger of duplicate records, a corrected identity number, or a later list that is itself incomplete. The 2025 comparisons count absences; the ministry posts cited here document additions with stated reasons, and none of them gives a count of records removed.

Matching names across rounds also depends on fields that do not stay fixed. The ministry’s list to 30 April 2024 drops both the date of birth and the source column and adds an address column, while the lists to 29 March and 30 June 2024 carry no address. The identity column is headed الهوية (identity) in March and ر.الوثيقة (document number) in June and August 2024. A record compared across rounds is therefore compared across schemas, and which field anchors the match changes what counts as the same person.

Which ministry totals carry the wrong date?

The Ministry of Health in Gaza’s own split for 30 June 2024 sums to 37,900, which is the total in its daily report of 1 July 2024 (message 5578), not the 37,877 its 30 June report gave (message 5576).

Both are the ministry’s own counts, published by a party to the war and not independently verified, and the mismatch sits entirely inside its own documents.

The same one-day offset recurs in the ministry’s national reporting. The Health Annual Report, Palestine 2024 gives 45,553 deaths and 108,379 injuries for the period from 7 October 2023 to 31 December 2024. Those are exactly the figures in the daily report of 1 January 2025 (message 6197). The ministry’s 30 December 2024 report gave 45,541, and no report was issued on 31 December. In both cases the period-end figure matches the next morning’s report.

Source: Ministry of Health, Gaza (2024 to 2026), messages 5576, 5578, 5651, 6194 and 6197, and Health Annual Report, Palestine 2024, section 1.1; a party’s own counts, not independently verified.

The offset has a measurable consequence. Subtracting the 28,185 complete records from the 30 June total of 37,877 gives 9,692, which is the unidentified figure Jamaluddine and colleagues (2025) report for that update, while the ministry’s own stated count of incomplete records is 9,715. A reader using the ministry’s 30 June total and a reader using its 30 June split arrive at two different backlogs from the same day’s publications.

The 2024 annual report adds a denominator slip of a different kind. It states 13,319 children, 7,216 women and 3,447 elderly people “with complete data till October 7, 2024,” beside a complete-data total of 43,587 at 11 December 2024 and 45,553 at year end. Any share computed from those figures divides a numerator from October by a base from December. The three category counts also do not partition anything: they sum to 23,982, the report publishes no count for men, and its 1,057 healthcare workers overlap all three categories.

Why do some birth dates on the list read as future years?

In the Ministry of Health in Gaza’s list to 31 August 2024, its own record of 34,344 deaths across 649 pages and not independently verified, the oldest decedents carry 2-digit birth years: a birth date printed 01-01-27 sits beside an age of 97.

Software that reads 2-digit years as this century renders that as 2027, a birth date after the death.

The Henry Jackson Society, a London policy institute, published a report in December 2024, “Questionable Counting,” that described age errors in the ministry’s lists, and it is often said that the lists contain impossible ages. The published fields show one documented way an apparent age error is produced. The same August list carries 01-01-24 beside an age of 100 and 01-01-22 beside an age of 101, which a naive reading turns into births in 2024 and 2022.

Source: Ministry of Health, Gaza (2024 to 2026), list of martyrs with complete data to 31 August 2024; a party to the war reporting its own count, not independently verified.

The formats also differ by source within a single document. In the ministry’s list to 29 March 2024, rows labelled as ministry records print full dates with slashes, in the form 30/03/1993, while rows labelled as notified by relatives print 2-digit years with dashes, in the form 13-02-97. A parser built for one format misreads the other, and the misreading is concentrated in one provenance stratum.

The age column is the check that the published fields supply. Read against the printed age, 01-01-27 can only be 1927, and the record is internally consistent. Read without it, the same row produces a decedent born in the future. The list to 30 April 2024 carries no date of birth at all, so the check works only on the rounds whose schema includes both fields.

The ministry’s own statement of 29 July 2024 bears on the ages that remain. Because dates of death were recorded on a family’s report or on arrival at hospital, an age computed at registration can exceed the age at death, which is the error the ministry says its audit committee found among decedents under 18. Both mechanisms, a 2-digit year misread downstream and a death dated late upstream, produce an age that does not match a person’s life, and only the second originates in the ministry’s own recording.

How many listed deaths rest on a relative’s report?

Of the 28,185 records in the Ministry of Health in Gaza’s list to 30 June 2024, 5,817 carry the label تبليغ ذوي الشهداء, notification by the martyrs’ relatives, rather than ministry hospital records. The label is the ministry’s own provenance field, and the count is a party’s own, not independently verified.

That is 20.6% of the June list. In the list to 29 March 2024, 2,478 of 21,323 records, or 11.6%, carried the relatives’ label, and the other 18,845 were labelled ministry records. Both partitions sum exactly to their list totals, so the source column accounts for every record in each document.

The base rate runs the other way from the premise. On the ministry’s own labels, 22,368 of the 28,185 June records, or 79.4%, come from its hospital and morgue records. The relatives’ stratum is the minority, though it grew between March and June 2024 as a share of the list.

Jamaluddine and colleagues (2025) identify what the relatives’ stratum is. The study describes its 2 ministry input lists for the identical window as 22,368 identified decedents from hospital morgue records and 5,817 individual decedents from the ministry’s rolling online mortality survey, launched on 1 January 2024, through which families report deaths. Those are precisely the 2 strata of the ministry’s 30 June list.

A relative’s report is a different evidentiary basis from a body received at a morgue, and the ministry flags it as such rather than merging it. The difference matters for the fields as well as the count: the 2-digit, dash-separated birth years in the March list sit in the relatives’ rows, and the ministry’s 29 July 2024 statement names the family’s report as one of the 2 ways a date of death was recorded. The column that marks a record as resting on a relative’s word is the same column that predicts which fields are weakest.

Did the peer-reviewed test find too many deaths or too few?

Jamaluddine and colleagues (2025), a peer-reviewed capture-recapture study in The Lancet, estimated 64,260 deaths from traumatic injury in Gaza between 7 October 2023 and 30 June 2024 (95% CI 55,298 to 78,525). The Gaza Ministry of Health, a party to the war reporting its own count, had reported 37,877 by the same date.

Capture-recapture estimates a total from the overlap between independent lists: the fewer names that appear on more than one list, the more deaths no list has caught. The study combined 3 lists, the ministry’s hospital morgue records, its online mortality survey, and obituary pages gathered from social media, and averaged alternative models over their overlap. The finding is too few, not too many, as set out in whether the Gaza health ministry undercounts or overcounts deaths.

The study’s headline 41% is a share of its own estimate. The gap between 64,260 and 37,877 is 26,383 deaths, which is 41% of 64,260 and 70% of 37,877; the ministry count was not low “by 41%” in the sense of multiplying 37,877 by 1.41. The interval is wide, and its lower bound of 55,298 still exceeds the ministry’s reported total.

Three limits travel with the estimate. It counts traumatic injury only, excluding deaths from disease, food insecurity and the collapse of health care. It excluded the 9,692 unidentified decedents in the ministry’s 30 June update, 26% of the ministry’s cumulative total, along with 930 people reported missing from the survey list. And 2 of its 3 input lists are the 22,368 and 5,817 strata of the ministry’s own 30 June document, so the study is an academic estimate built on the ministry’s records, not a count of the dead made separately from them.

Within those limits, the test bears on the direction of error. Duplicates, misdated deaths or wrongly attributed records would inflate a list; missed deaths deflate it. The single peer-reviewed measurement places the true traumatic-injury toll above the ministry’s figure, with errors of both kinds possible inside that net result.

How far did Israel’s own 2014 count differ from the ministry’s?

The Gaza ministry’s count of deaths in the 2014 war differed from the Israeli Foreign Ministry’s by 8.0%, according to a Lancet correspondence letter by Huynh, Chin and Spiegel (2024), which was editorially reviewed and not peer-reviewed. Both underlying figures are a party’s own count, relayed by an Associated Press explainer by Isabel Debre of 26 October 2023.

The pair AP printed is 2,310 deaths on the Gaza ministry’s count and 2,125 on Israel’s, a difference of 185. That is 8.0% of the Gaza ministry’s figure and 8.7% of Israel’s, and only the first reproduces the letter’s number. In the one comparison on record against Israel’s own published count, the Gaza ministry’s total was the higher of the two.

The letter’s companion range, a discrepancy of 1.5% to 3.8% between ministry reporting and United Nations analyses in previous conflicts, rests on a sentence that carries no citation marker in either the journal’s web text or its typeset PDF. The range reproduces from the 2021 and 2008 pairs in the same AP explainer, again only with the ministry’s count as denominator; AP itself prints no percentage. The UN figures in those pairs are post-war tolls of the UN humanitarian office, OCHA, and AP’s pairs omit the November 2012 conflict, so “previous conflicts” covers 3 wars, not every one.

The letter’s own measurement is narrower. It compared cumulative reported deaths over 35 days, 7 October to 10 November 2023, finding 11,078 Gazans and 101 UNRWA staff, and wrote that the ministry’s rate of 5.3 deaths per 1,000 residents was lower than the 7.8 per 1,000 among UNRWA staff. Those are rates over 2 different populations. The authors added that ministry mortality “might have been under-reported” in the period, and their title, “No evidence of inflated mortality reporting,” is an absence claim rather than a verification.

A search of the biomedical and preprint literature on 10 September 2026 located no peer-reviewed study comparing a pre-2023 ministry count with any other body’s count. The comparison circulates as a percentage; as a study, it has not been carried out.

Which errors enter after other bodies relay the count?

Crawford (2025), a Costs of War report from Brown University that is not peer-reviewed, gives 67,075 killed in Gaza to 3 October 2025. The figure is the Gaza Ministry of Health’s own count, not independently verified, taken from a United Nations dashboard, so the chain runs from ministry to dashboard to paper, and an error at any link is not the list’s.

Crawford adds the ministry’s 67,075 killed and 169,430 injured to reach 236,505 casualties, and reports that sum as more than 10% of Gaza’s pre-war population of about 2.2 million. The 10% is a share of casualties, killed and injured together. Restated as a share of Gaza’s population killed, it overstates the ministry’s own death count more than threefold, an error produced in the relay, not in the list. The paper separately reports a later population estimate of 2.1 million, and the 2 denominators are not interchangeable.

The paper’s civilian share of about 80% does not come from the ministry, which publishes no combatant breakdown. Crawford’s own footnote derives it from a Guardian report of Israeli military data through May 2025 and flags the assumption, so a civilian proportion attributed to the ministry’s lists cites a field the lists do not contain.

Relays of the peer-reviewed estimate drift the same way. The “more than 70,000” figure attached to Jamaluddine and colleagues (2025) in press coverage is not their capture-recapture result. It is the authors’ own extrapolation in the paper’s Research in context box, applying the same under-reporting rate to the ministry’s October 2024 figure of 41,909, worded “probably exceeded 70 000.” Their 41%, a share of the 64,260 estimate, is also read as a multiplier on the ministry’s 37,877.

The Lancet letter by Huynh, Chin and Spiegel (2024) undergoes the same drift. Its title records an absence of evidence of inflation, and its text allows for under-reporting; restated as a finding that the ministry’s count was verified, it reverses the direction of what the letter wrote. Each of these is a documented misreading that enters after the ministry publishes, attributable to the step that introduced it.

Methodology and limitations

The evidence draws on 3 source types. The primary record is the Gaza Ministry of Health’s own publication: 4 name-by-name list documents covering deaths to 29 March, 30 April, 30 June and 31 August 2024; daily statistical reports at the dates cited; the ministry’s statements of 7 January, 24 July and 29 July 2024 and its additions of 2025 and 2026; and the Health Annual Report, Palestine 2024 and 2025. All were read on 10 September 2026. Every ministry figure is a party’s own count and none is independently verified.

The peer-reviewed tier is Jamaluddine and colleagues (2025), a 3-list capture-recapture study whose interval is wide and which counts traumatic-injury deaths only. Huynh, Chin and Spiegel (2024) is Lancet correspondence, editorially reviewed and not peer-reviewed, and its pre-2023 range carries no citation. Crawford (2025) is a university research project report, not peer-reviewed, compiling counts it did not measure. Press and policy-institute publications, The Jewish Chronicle, HonestReporting through The Jerusalem Post, the Foundation for Defense of Democracies, the BBC and the Associated Press, are named as the holders or relays of specific claims and dated, and their figures are carried as those publishers report them.

Retrieval shapes what is checkable. The ministry’s website at moh.gov.ps holds 2,536 files, the newest dated 21 October 2023 and none a casualty list; the lists are posted only to the ministry media office’s Telegram channel, which serves message text but no files. The list documents were therefore read from a public mirror of the posted files and matched against the ministry’s own stated record counts, page counts and column sets, and for the March file against its stated file size. Ages and birth dates were read against each other row by row rather than parsed mechanically. Records are reported in aggregate, and no decedent is named.

Conclusion

So have errors been found in the fatality lists issued by Gaza’s Hamas-run health ministry? Yes, and the first body on record documenting some of them is the ministry itself, a party to the war whose counts are not independently verified, through its 29 July 2024 statement that an audit found decedents registered months after death.

The documented errors are specific and mixed in direction. The ministry’s own 30 June 2024 split sums to 37,900, its 1 July total, and its 2024 annual year-end figure is the next morning’s report. Its August 2024 list prints birth years like 01-01-27 that only the age column rescues. 5,817 of its June records rest on a relative’s report, by its own label. The Jewish Chronicle and HonestReporting counted about 3,400 names absent from a 2025 round, while the ministry’s own channel records committee additions of 233, 280 and 160. Against all of that, the only peer-reviewed test put traumatic-injury deaths at 64,260, above the ministry’s own 37,877, and Israel’s own 2014 count sat 8.0% below the ministry’s. Several of the most repeated errors, the 10% read as deaths, the 41% read as a multiplier, enter only after other bodies relay the figures.

A list that is corrected in public, round after round, by the party that compiles it, leaves a trail of its own revisions, and a toll that no one corrects leaves none. When the errors in a count of the dead are documented in both directions at once, and the one statistical test places the total higher, what would it take for the argument over the lists to turn into a ledger of their changes?

Frequently Asked Questions

Does the ministry’s list record a date of death for each person?

No. The 4 list documents read for 2024 carry serial number, name, an identity or document number, date of birth, sex, age and in most rounds a source column, and the phrase تاريخ الاستشهاد (date of martyrdom) appears in no column header (Ministry of Health, Gaza, 2024 to 2026). The ministry’s own 29 July 2024 statement explains that dates of death were recorded on a family’s report or on arrival at hospital.

What does the ministry say its count leaves out?

On 7 January 2024 (message 4739) the Gaza ministry, reporting its own count, listed 5 exclusions: the missing and those under rubble, the unidentified who reached hospitals, the unidentified bodies Israel returned, families’ own burials outside hospitals, and deaths in Gaza and North Gaza after its information system stopped in November 2023. The statement has not been withdrawn in the ministry’s published channel.

Has the ministry revised a single-incident death toll?

The BBC reported in 2023 that the ministry first gave 500 deaths for the 17 October 2023 explosion at Al-Ahli hospital and later revised its own figure to 471. The BBC also reported a US intelligence estimate of 100 to 300 deaths, likely at the low end. That revision concerns one incident’s count, not the named lists.

Where does the ministry publish its lists?

The ministry’s own website has published no file since 21 October 2023, so its lists appear only on its media office’s Telegram channel, with corroborating national reports on site.moh.ps (Ministry of Health, Gaza, 2024 to 2026). The interactive platform it now names, sehatty.ps, returned an error on the path the ministry publishes when checked on 10 September 2026.

How does the ministry break the dead down by age and sex?

The ministry’s Director General stated on 23 March 2025 that 15,613 of 50,021 dead were children, the ministry’s own count, and it publishes separate lists for children, women, the elderly and orphans alongside each combined list. Its Health Annual Report, Palestine 2025 gives children 30%, women 16%, the elderly 7% and men 47% of 71,271 deaths to the end of 2025, again its own figures.

Has the count kept rising since the ceasefire?

The ministry’s daily report of 10 September 2026 gave 73,670 cumulative deaths, its own count, including 1,359 since the ceasefire of 11 October and 831 bodies recovered from rubble. Recovered bodies enter the total as they are found, so part of any post-ceasefire rise reflects deaths that occurred earlier.

Sources

• Associated Press (Isabel Debre), 2023. What is Gaza’s Ministry of Health and how does it calculate the war’s death toll? Published 26 October 2023, updated 6 November 2023. apnews.com, article identifier death-toll-59470820308b31f1faf73c703400b033. International press.
• BBC, 2023. How the dead are counted in Gaza. https://bbc.com/news/world-middle-east-67347201. International press.
• Crawford, Neta C., 2025. The Human Toll of the Gaza War: Direct and Indirect Death from 7 October 2023 to 3 October 2025. Costs of War project, Watson School of International and Public Affairs, Brown University. https://costsofwar.watson.brown.edu/paper/HumanTollGaza. University research project report, not peer-reviewed.
• Foundation for Defense of Democracies, 2024. Hamas-run Gaza Health Ministry admits to flaws in casualty data. 9 April 2024. https://www.fdd.org/analysis/2024/04/09/hamas-run-gaza-health-ministry-admits-to-flaws-in-casualty-data/. Policy institute analysis.
• The Guardian, 2025. Report of Israeli military data on Gaza fatalities through May 2025, as cited in Crawford (2025). International press.
• Huynh, Benjamin Q., Elizabeth T. Chin and Paul B. Spiegel, 2024. No evidence of inflated mortality reporting from the Gaza Ministry of Health. The Lancet, 403(10421), pp. 23 to 24; published online 6 December 2023. DOI 10.1016/S0140-6736(23)02713-7, PMID 38070526. Correspondence, editorially reviewed, not peer-reviewed.
• Jamaluddine, Zeina, Hanan Abukmail, Sarah Aly, Oona M. R. Campbell and Francesco Checchi, 2025. Traumatic injury mortality in the Gaza Strip from Oct 7, 2023, to June 30, 2024: a capture-recapture analysis. The Lancet, 405(10477), pp. 469 to 477. DOI 10.1016/S0140-6736(24)02678-3, PMID 39799952. Peer-reviewed.
• The Jerusalem Post, 2025. HonestReporting analysis of absences from the Gaza Ministry of Health’s March 2025 fatality list. https://www.jpost.com/israel-news/defense-news/article-848592. International press, reporting a media-monitoring organization’s analysis.
• The Jewish Chronicle, 2025. Hamas death toll figures are a sham, and now we have proof. https://www.thejc.com/opinion/analysis/hamas-death-toll-figures-are-a-sham-and-now-we-have-proof-tym87dsn. Press analysis.
• Ministry of Health, Gaza, 2024 to 2026. Lists of martyrs with complete data to 29 March, 30 April, 30 June and 31 August 2024; daily statistical reports; statements of 7 January, 24 July and 29 July 2024, 23 March and 30 August 2025, and August 2026. Ministry media office channel, https://t.me/MOHMediaGaza, messages 4739 to 7607. Official or primary record, a party to the war reporting its own count, not independently verified.
• Ministry of Health, Palestine, 2025 and 2026. Health Annual Report, Palestine 2024; Health Annual Report, Palestine 2025. https://site.moh.ps/index/Books/BookType/2/Language/ar. Official or primary record, a party’s own count, not independently verified.