352,000 deaths in four years. Mediazona and Meduza’s new estimate of Russian losses in Ukraine

Our methodology draws on data from the Probate Registry—a publicly accessible Russian database of inheritance cases—and the recorded names count we maintain together with BBC News Russian and a team of volunteers (now at 217,808 verified names). We described it in earlier articles, but here is a brief overview.

The Probate Registry includes every deceased citizen who left significant assets to be inherited. By cross-referencing it against our named list of losses, we can calculate the probability of inclusion in the registry for each age and social group. That coefficient then lets us “convert” inheritance cases into an actual losses figure—that is, to answer the question: how many real deaths does the known number of inheritance cases correspond to?

The estimate therefore captures everyone, not only those who left an inheritance. For groups unlikely to appear in the registry (prisoners, for example) we apply a higher conversion coefficient, and they too are included in the overall total.

As before, the estimate covers only Russian citizens whose deaths have been officially confirmed. But this update is the first in which we have had to account for an entirely new category of casualties: those whose deaths have been confirmed by the courts... without a body.

Of the 352,000, some 261,000 are “regular” fatalities, calculated by the same method as in our previous estimates. The remaining 90,000 or so are men declared dead or missing by court order, or whose deaths were registered with a delay of at least 180 days for other reasons: body exchanges, say, or late identification.

We can speak about “regular” and “late” fatalities with differing degrees of precision. We know more about the first group, which lends itself more readily to analysis. We know less about the social composition of the second. Most importantly, the figure of 90,000 “late” fatalities is an estimate as of mid-2025: more recent late fatalities are simply not yet visible due to the same delay. For at least 180 days after their actual death, these soldiers do not appear either in the courts or in the Probate Registry, so more time will need to pass before we can estimate the late category through to the end of 2025.

It is important to stress that this category refers to people who died in comba, not those captured or who fled the battlefield. They may at some earlier point have been listed as missing in action, but the 90,000 figure refers to those already officially recognised as dead.

It would be incorrect to say that we have started counting missing personnel separately: the only difference between the two categories lies in the mechanism by which a death has been officially registered. We do not currently know how many service members are formally listed as missing.

The “court-ordered” or “late” deaths are clearly visible both in the Probate Registry and on court websites. In the registry, the group stands out for the long gap between the date of death and the date of its registration. These are people who died and then went 180 days or more without a death certificate being issued: either the body was not recovered immediately, or it was not identified, or it was never returned at all (the last being the typical case for a court-ordered declaration of death or missing status).

Court applications to declare service members missing or dead began to be filed en masse in July 2024, as Mediazona has reported in detail. Data on the applicants confirms that these claims concern military personnel: applications relating to civilians have held steady at around 8,000 a year.

At the same time, “late” death registrations have become noticeably more common in the Probate Registry, where they were previously rare. Since 2024, we have identified 52,000 excess probate cases with death registrations more than 180 days after the date of death. Applying the average rate at which probate cases convert to actual deaths yields 90,000 fatalities.

A similar number emerges from the court data. In total, 86,500 “excess” claims (that is, exceeding the pre-war level) were filed in 2024 and 2025.

In theory, such claims can be duplicated, with more than one filing for the same person. This occurs when someone is first declared missing and later declared dead, or when a judge returns a claim and the military unit refiles it with corrections.

A conservative lower estimate of the number of unique claims is 61,800. To arrive at that lower bound, we subtracted every category that could in principle generate duplicates: claims that were returned, dismissed or left without consideration, as well as claims seeking to have a person declared missing rather than dead, since they could subsequently be the subject of a separate ruling declaring the same person dead.

To isolate that second category, we reviewed 3,800 rulings on missing service members, supplied by a source in the judicial system. Of these, 2,600 claims sought to have a person declared dead and 500 sought to have them declared missing. The remaining 700 were procedural decisions: error corrections, refusals, resubmissions.

Court claims clearly do not account for the entire body of “ultra-late” registrations visible in the Probate Registry. Although such cases were almost nonexistent before the wave of court-led declarations (they make up just 2.4% of early snapshots of the named list), their number could have grown independently of the courts, for instance through more frequent exchanges of bodies.

Furthermore, the court data only takes us up to mid-2025: although the applications themselves run to December, by law they concern people who went missing at least 180 days before the filing date.

Taking this into account, and given that applying the “regular” conversion coefficient to inheritance cases with “ultra-late” registration produces the same figure as the non-conservative count from the courts (90,000 from the coefficient against 86,500 from the claims), we have decided to estimate this group at the upper end of the court-claims range: 90,000 people.