A household survey puts Gaza's violent death toll a third above the official count

For more than two years, the argument over how many people have died in the Gaza Strip has been an argument about a single set of numbers: the running violent death toll published by the Gaza Ministry of Health, which has drawn both sustained criticism and sustained defence [s1]. A study published in The Lancet Global Health this week does something different. Rather than auditing the ministry's list, researchers went and counted [s1].

What was done

Between 30 December 2024 and 5 January 2025, investigators conducted a population-representative household survey — the Gaza Mortality Survey — covering 2,000 households across 200 primary sampling units [s1]. They documented the vital status of 9,729 people who had been household members as of 6 October 2023, the day before the conflict began, plus children born since [s1]. The sample was stratified by dwelling type and by governorate of origin, and the analysis used raking to adjust for demographic composition, with confidence intervals calculated by Taylor series linearisation [s1].

That last detail matters more than it sounds. Sampling could only occur in areas that were physically accessible, so households displaced out of inaccessible governorates stand in for the populations of those places [s1]. This is a standard and unavoidable compromise in conflict mortality work, and it is one of the reasons a survey of this kind produces an estimate with a stated range rather than a count.

What it found

The survey estimated 75,200 violent deaths (95% CI 63,600–86,800) between 7 October 2023 and 5 January 2025 — roughly 3.4% of the Gaza Strip's pre-conflict population [s1].

The Ministry of Health figure covering the same period was 49,090 violent deaths, which sits 34.7% below the survey's central estimate [s1]. In other words, the survey's finding runs against the direction of the most common criticism levelled at the ministry's reporting: it suggests the official toll has understated, not overstated, violent deaths.

On the demographic composition of the dead, the two sources converge. The survey estimated that women, children under 18 and people over 64 made up 56.2% (95% CI 50.4–61.9) of violent deaths, totalling 42,200 deaths (95% CI 33,100–51,300) — a distribution the authors describe as aligning with ministry reporting [s1].

The survey also produced the first independent estimate of deaths from causes other than direct violence: 16,300 non-violent deaths (12,300–20,200), of which 8,540 (4,540–12,500) were excess deaths above what pre-conflict trends would have predicted [s1]. Those excess non-violent deaths are substantial, but the authors note they are lower than some projections had suggested [s1]. Both halves of that sentence deserve equal weight.

The uncertainty is the point

The confidence intervals here are wide by the standards of peacetime epidemiology, and they should be read as part of the finding rather than as a footnote to it. The violent-death interval spans roughly 23,000 deaths; the excess non-violent interval spans nearly 8,000 — enough that the lower bound and the upper bound imply materially different pictures of how the health system's collapse has translated into deaths.

The journal published an accompanying comment alongside the study, framed explicitly around what the Gaza Mortality Survey reveals and what it misses about counting deaths in the Gaza Strip [s2]. That framing is a fair summary of where the evidence now sits: a single cross-sectional survey, conducted over one week, in accessible areas, with displaced households representing the areas that could not be reached [s1].

How it fits with other recent estimates

The survey is not the only quantitative work to come out of this period. A separate analysis published in the European Journal of Public Health in January used official data sources cross-referenced against professional-association lists to estimate mortality among two groups protected under international humanitarian law — healthcare workers and journalists — relative to Gaza residents of the same age and sex [s3]. It found consistently higher mortality risk in both groups, ranging from 36% higher to more than sixfold higher for journalists [s3].

These are different questions answered with different methods: one estimates a population total from a sample, the other estimates relative risk in named occupational groups from list-based data. What they share is an attempt to replace inference-by-assertion with something that carries an explicit method and an explicit error term.

What to watch

The survey's authors argue that their work demonstrates the feasibility of mortality surveillance inside an active conflict zone [s1]. If that holds, the more useful question going forward is not whether a single number is right but whether repeated surveys can track the trajectory — particularly for non-violent deaths, where the mechanism is a degraded health system rather than direct injury, and where the effect accumulates slowly.

The study was funded by a European Research Council grant and the Centre for Research on the Epidemiology of Disasters at Université Catholique de Louvain [s1]. An Arabic translation of the abstract was published with it [s1].

Sources

• [s1] Violent and non-violent death tolls for the Gaza conflict: new primary evidence from a population-representative field survey. The Lancet Global Health, 18 February 2026. https://doi.org/10.1016/S2214-109X(25)00522-4
• [s2] From enumeration to inference: what the Gaza Mortality Survey reveals — and misses — about counting deaths in the Gaza Strip. The Lancet Global Health, 18 February 2026. https://doi.org/10.1016/S2214-109X(26)00015-X
• [s3] Mortality risk for healthcare workers and journalists in the Gaza Strip over 2023–24. European Journal of Public Health, 13 January 2026. https://doi.org/10.1093/eurpub/ckaf241