How many Russian soldiers died - according to regional budgets?

This post presents a new method for estimating the number of Russian fatalities based on budget spending in certain Russian regions. The results appear to match the data of other OSINT projects.

The number of Russian soldiers killed¹ in the so-called “special military operation” is one of the Kremlin’s best-kept secrets. This is to be expected, given the political sensitivity of this statistic. There are also wildly different numbers circulating in the international political discourse. In recent months, Donald Trump has repeatedly cited figures that appear to be far too high. Adding to the confusion, others have repeated these numbers, either out of carelessness, political convenience, or confusion between casualty and fatality statistics.

At the same time, it is my impression that there is actually a general consensus among long-standing observers of the war on the magnitude of Russian war fatalities. Over the years, several methods have been developed to estimate the death toll among soldiers based on publicly available data, and most of these methods would put the total figure somewhere between 200,000 and 300,000 (as of autumn 2025). A few of these projects are:

• Mediazona/BBC’s 200: Based on obituaries and other publicly available sources collected by volunteers, the project lists the identities of 135,100 soldiers who have died since 2022 (as of October 10, 2025). While the method only captures a subset of all KIAs, it is highly detailed and lends credence to total KIA estimates above 200,000. According to early estimates, the list may cover 40%-60% of the total.

• iStories’ Charon - a project that uses AI and volunteer verification to collect the identities of killed and missing soldiers, similar to, but independent from Mediazona. As of July 11, 2025, the project contains the names of 117,978 Russian soldiers who were killed or went missing. Most likely, a similar “multiplier” has to be applied as for “200”.

• Meduza’s probate registry estimates (quoted in Mediazona’s 200: 219,000). The unusually high number of inheritance cases involving younger men after 2022 makes it possible to calculate the number of inheritance cases linked to war-related deaths.

• Before it was censored, official Russian demographic data made it possible to estimate war-related deaths in a similar way (the method in short: young male excess mortality). See, e.g.: Kobak et al. 2025 in Demography, which puts the lower bound of war fatalities in 2022-2023 at 58,500 (it doesn’t include certain categories of KIA).

• A completely different approach is pursued by Andrew Perpetua, who has been counting killed Russian soldiers based on combat footage (a very useful, but certainly very grim task). He visually confirmed 34,943 Russian KIA in footage covering 289 days (121 per day), and he estimates that the true figure is more than twice that (up to 102,902 per year).

I would like to introduce another independent data source that can be used to estimate Russian war fatalities, albeit only for a few Russian regions: regional budget data. The results closely align with data collected by Mediazona on a regional level, although it may not necessarily represent the same underlying cases, as I will explain below.

Most Russian regions offer payouts to the families of deceased (and sometimes: missing) soldiers. These regional payouts typically range from 1 to 3 million rubles, which is only a small part of the families’ total compensation package. This package usually also includes the presidential payout (Ukaz 98, 5 million rubles), the federal payout (Federal Law 306, 4.8 million rubles), and mandatory life insurance (3.44 million rubles).

Unlike regional sign-on bonuses, the size of regional payouts to families has not changed since 2022 in the observed regions (most likely because they are not that important for recruitment success, so there is no reason to increase them). This makes it easier to estimate the total number of cases in which compensation was paid: Total budget spent divided by the size of the regional payout.²

It’s important to note that this data comes from a completely different source not related to Mediazona’s project 200. It’s based solely on information published by the Russian government (mostly: the Finance Ministry). Still, the regional patterns appear to be almost identical. Unfortunately, only very few Russian regions offer suitable information, but the results are appear to be conclusive. As with all other approaches, this is not the “true KIA count,” but rather another incomplete piece of the puzzle.

Budget spending on payouts to families accelerated sigfnificantly in the second half of 2024. One reason for this could be the changing composition of killed Russians: Wagner merceneries and prisoners without a MoD contract, who accounted for a significant share of the death toll in 2022 and 2023 (according to Mediazona), would in most cases not qualify for a regional payout, and possibly get compensation from Wagner directly.³

Although the specifics of who received which compensation from regional budgets are somewhat uncertain, a more appropriate comparison could be to contrast non-mercenary, non-prisoner deaths from Mediazona with the results from regional budget data. That comparison would yield a significantly higher death toll from regional data, indicating many additional cases not included in the Mediazona database. Or not included yet: It is unclear whether cases generally appear earlier in the Mediazona list or in the budget data. I would assume the latter, given the rather low totals on Mediazona for the last six months. A time lag between both data sources would also affect the comparison.

The twelve regions considered here have a combined population of 26.3 million, which is around 18% of Russia’s total population, and the total death toll is 23,600. On average, the number of fatalities is close to 0.1% of the total population in these regions. It varies from 0.16% in Bashkortostan and 0.04% in Moscow Oblast. Extrapolating from such a small population sample will not yield robust totals for Russia as a whole, but the results would also not contradict the totals of other OSINT projects.

These calculations are an initial attempt to estimate Russian war fatalities using regional budget data, so exercise caution when interpreting the results. This is also a new topic for me on this blog, so I apologize for any mistakes I may have made (I may update this post a few times). If you have any suggestions for improvement, criticism, or questions, please let me know.

1. I’m aware that the terminology surrounding Russian war fatalities is complex, even if civilian deaths are not included. Most Russian soldiers died in combat in Ukraine (KIA). However, some died for reasons indirectly related to the war, such as training accidents, or they died in Russia. This post simplifies, hopefully it doesn’t oversimplify. For the estimate based on regional budget data, it is relevant whether the death of a soldier entitles his relatives to a payout. This should also be the case for some non-combat deaths, at least if they occur in Ukraine.

2. I could not find any public information about the size of the payout in Tver Oblast. However, it is likely to be 1 million rubles. Almost all neighboring regions offer 1 million rubles. The budget line for payouts to families in Tver appeared in March 2022, and the initial payment was 1 million rubles on March 17, 2022 (indicating a short time lag). At the same time, a second budget line appeared for funeral costs. I also cross-checked the total payouts to families with the total payouts for soldiers’ funerals and gravestones in Tver. The results of both methods were close (it is more difficult to compare because funeral and gravestone payouts changed more often).

3. The increase in payouts in 2024 could also be related to instructions by Vladimir Putin (effective 1 April 2024) that require authorities to pay compensation within three months and allow residents of other regions to receive regional payouts for deceased relatives.