The government of Venezuela says the death toll from the twin earthquakes was 6,509. But a New York Times investigation, based on estimates from the U.S. Geological Survey (USGS) and the Times’ own review of satellite imagery and housing records, concludes that the official figure may be a severe undercount.
In stark contrast to the official figure, the USGS puts its median estimate at about 53,000 deaths. The agency also estimates an 81% chance that the true toll exceeds 10,000. The government has not updated its count since Aug. 24.
How the USGS arrived at its earthquake death toll estimate in Venezuela
For years, the USGS has estimated earthquake deaths by combining shaking severity, historical data and population distribution. David Wald, a USGS seismologist, told the Times that such estimates tend to be more reliable in high-casualty events like this one. The 53,000 figure is a median, meaning the agency sees an equal chance that the actual number is higher or lower.
A major complication is the lack of an up-to-date official census. To work around this, the USGS uses a model called LandScan, developed at Oak Ridge National Laboratory, which projects the most recent census data forward in time. The model draws on United Nations figures and dozens of other variables, including land use, roads and proximity to cities, to estimate where people were when the ground shook.
Inside the collapsed housing towers
The New York Times investigation into the death toll in Venezuela focused on public housing in La Guaira state, the hardest-hit area. Its analysis found that 16 concrete towers collapsed, each with 12 residential floors and eight apartments per floor, according to floor plans and site observations. Reporters estimated an average of 3.4 people per apartment based on a community council census of one complex. They corroborated that figure with former residents and a government statement that put the number at about four people per apartment.
That works out to roughly 5,200 people living in those 16 towers alone, close to 80% of the government’s nationwide death toll. Not everyone would have been home when the earthquakes struck. But witness accounts and surveillance footage indicate that at least some buildings collapsed immediately, leaving people inside little or no time to escape. Times reporters on the ground saw bodies being recovered from some buildings every day for weeks.
Garces oversees a government commission reviewing the damage. His figure, given to the Times, is more than triple the 190 buildings the government has publicized as collapsed. Copernicus, the European Union’s Earth observation program, found nearly 700 structures destroyed and hundreds more damaged.
The missing and the uncounted
Venezuelan officials have refused to say how many people are still missing and presumed dead, a major source of anger among survivors. In the absence of official reports, Venezuelans have started compiling their own lists. Jorge Bastidas, a programmer who created the Venezuela Earthquake Disappeared website, said locals have reported more than 29,000 missing people there. He acknowledged that some entries are likely duplicates or refer to people who have since been found, despite his team’s efforts to clean the list. However, he said the figure still points to thousands of potentially uncounted deaths. A second effort, Venezuela Reports, lists more than 40,000 missing people and has also tried to remove duplicates.
Why an accurate count is so hard
Two obstacles stand out in the New York Times’ investigation to determine the real death toll in Venezuela. First, there is no recent census. Venezuela has not conducted a rigorous field census since 2011, according to Robert Rodriguez, a Venezuelan researcher and former election official.
To estimate how many people were in the hardest-hit part of La Guaira, he tallied registered voters in the seven worst-hit precincts and added 38% for children, visitors, and unregistered voters. His estimate is 342,000 people present on the day of the quakes.
There is also no official update on the dead or missing. The government’s last public count came two months after the quakes, and it has not said how many remain missing. That leaves families, researchers, and outside analysts to piece the picture together from estimates and citizen-run lists.
Why the number matters
Oscar Murillo, who leads the Venezuelan human rights group Provea, said an accurate count is essential for both healing and accountability. It would help victims’ families feel seen, he said, and supply the information needed to decide whether criminal or civil investigations are warranted over deaths that could have been avoided.
The Times separately reported on how the public housing towers, known as OPPPEs after the government office that built them, were designed and constructed, and found flaws that proved deadly when the quakes hit. The towers became some of the densest pockets of death, the paper reported.
This report is based on an investigation by The New York Times published Oct. 21, 2026, by James Glanz, Julie Turkewitz and Helmuth Rosales, which drew on U.S. Geological Survey estimates, Copernicus satellite data and citizen-compiled lists. USGS figures are statistical estimates, not confirmed counts. Tables and charts were prepared from the figures reported there.