Deaths from bombs and other traumatic injuries during the first nine months of the war in Gaza may have been underestimated by more than 40 percent, according to a new analysis published in The Lancet.

The peer-reviewed statistical analysis, led by epidemiologists at the London School of Hygiene and Tropical Medicine, used modeling in an effort to provide an objective third-party estimate of casualties. The United Nations has relied on the figure from the Hamas-led Ministry of Health, which it says has been largely accurate, but which Israel criticizes as inflated.

But the new analysis suggests the Hamas health ministry tally is a significant undercount. The researchers concluded that the death toll from Israel’s aerial bombardment and military ground operation in Gaza between October 2023 and the end of June 2024 was about 64,300, rather than the 37,900 reported by the Palestinian Ministry of Health.

The estimate in the analysis corresponds to 2.9 percent of Gaza’s prewar population having been killed by traumatic injury, or one in 35 inhabitants. The analysis did not account for other war-related casualties such as deaths from malnutrition, water-borne illness or the breakdown of the health system as the conflict progressed.

The study found that 59 percent of the dead were women, children and people over the age of 65. It did not establish what share of the reported dead were combatants.

Mike Spagat, an expert on calculating casualties of war who was not involved in this research, said the new analysis convinced him that Gaza casualties were underestimated.

“This is a good piece of evidence that the real number is higher, probably substantially higher, than the Ministry of Health’s official numbers, higher than I had been thinking over the last few months,” said Dr. Spagat, who is a professor at Royal Holloway College at the University of London.

But the presentation of precise figures, such as a 41 percent underreported mortality, is less useful, he said, since the analysis actually shows the real total could be less than, or substantially more. “Quantitatively, it’s a lot more uncertain than I think comes out in the paper,” Dr. Spagat said.

The researchers said their estimate of 64,260 deaths from traumatic injury has a “confidence interval” between 55,298 and 78,525, which means the actual number of casualties is likely in that range.

If the estimated level of underreporting of deaths through June 2024 is extrapolated out to October 2024, the total Gazan casualty figure in the first year of the war would exceed 70,000.

“There is an importance to war injury deaths, because it speaks to the question of whether the campaign is proportional, whether it is, in fact, the case that sufficient provisions are made to to avoid civilian casualties,” said Francesco Checchi, an epidemiologist with an expertise in conflict and humanitarian crises and a professor at the London School of Hygiene and Tropical Medicine who was an author on the study. “I do think memorializing is important. There is inherent value in just trying to come up with the right number.”

The analysis uses a statistical method called capture-recapture analysis, which has been used to estimate casualties in other conflicts, including civil wars in Colombia and Sudan.

For Gaza, the researchers drew on three lists: The first is a register maintained by the Palestinian Ministry of Health, which mainly comprises the dead in hospital morgues and estimates of the number of unrecovered people buried in rubble. The second is deaths reported by family or community members through an online survey form the ministry established on Jan. 1, 2024, when the prewar death registration system had broken down. It asked Palestinians inside and outside Gaza to provide names, ages, national ID number and location of death for casualties. The third source was obituaries of people who died from injuries that were published on social media, which may not include all of the same biographical details and which the researchers compiled by hand.

The researchers analyzed these sources to look for individuals who appear on multiple lists of those killed. A high level of overlap would have suggested that few deaths were uncounted; the low amount they found suggested the opposite. The researchers used models to calculate the probability of each individual appearing on any of the three lists.

“Models enable us to actually estimate the number of people who have not been listed at all,” Dr. Checchi said. That, combined with the listed number, gave the analysts their total.

Patrick Ball, director of research at the Human Rights Data Analysis Group, and a statistician who has conducted similar estimates of violent deaths in conflicts in other regions, said the study was strong and well reasoned. But he cautioned that the authors may have underestimated the amount of uncertainty caused by the ongoing conflict.

The authors used different variations of mathematical models in their calculations, but Dr. Ball said that rather than presenting a single figure — 64,260 deaths — as the estimate, it may have been more appropriate to present the number of deaths as a range from 47,457 to 88,332 deaths, a span that encompasses all of the estimates produced by modeling the overlap among the three lists.

“It’s really hard to do this kind of thing in the middle of a conflict,” Dr. Ball said. “It takes time, and it takes access. I think you could say the range is larger, and that would be plausible.”

While Gaza had a strong death registration process before the war, it now has only limited function after the destruction of much of the health system. Deaths are uncounted when whole families are killed simultaneously, leaving no one to report, or when an unknown number of people die in the collapse of a large building; Gazans are increasingly buried near their homes without passing through a morgue, Dr. Checchi said.

The authors of the study acknowledged that some of those assumed dead may in fact be missing, most likely taken as prisoners in Israel.

Roni Caryn Rabin and Lauren Leatherby contributed reporting.