So, has Ukraine only lost 31,000 since the start of the war?

Well, Zelenskyy got my attention with his claim on Sunday that Ukraine has lost 31,000 killed in action since the start of the war. He also claimed that the Russians have lost 180,000. See: Zelenskyy says 31,000 Ukrainian soldiers killed since Russia invaded 2 years ago | CBC News).

Several issues here:

First, who makes up those 31,000 killed? If Zelenskyy has thrown out a figure for the first time since the summer of 2022, I am guessing it is based upon something. These figures probably do not include missing. Now we do not have a count of the missing. We do know that as of 30 June 2022 Russia claimed that there were at least 6,000 Ukrainians captured (see Battle for Kyiv, Chapter 20). In early July 2022, the Ukrainian missing person commissioner stated on TV that more than 7,000 people are missing, including soldiers, national guardsmen, border guards and intelligence officers. When on the defense and retreating, the missing can add up, sometimes exceeding the number reported as killed. This is shown by the Russian First Guards Tank Army reports for 24 February to 15 March 2022 (even though they were on the offiense), the missing outnumbered the killed (See: Tank Losses and Crew Casualties in the Russo-Ukrainian War | Mystics & Statistics ( and The Russian First Tank Army Report from 24 February – 15 March 2022 | Mystics & Statistics ( 

But, I am not sure that is all that is missing. Are Ukrainian National Guard included? The Ukrainian National Guard includes units such as the Azov Battalion. They are not organizationally part of the Ukrainian Army. Then there is the Territorial Defense Forces. These often ad hoc organizations are officially part of the Ukrainian Army, but not sure that their losses are being counted among the army losses. So, actually Ukrainian losses are clearly higher than 31,000.

The last time Zelenskyy gave an estimate was on 21 August of 2022. He said at the time that their losses were around 9,000. Two months earlier (11 June), there were reports that Ukraine was saying that their losses were 10,000. So they have a 1,000 less losses two months later? A presidential aide put Ukrainian losses as of 1 December 2022 at 10,000 – 13,000. So, we have 13,000 losses in the first 9 months of the war, and another 18,000 losses in the last 15 months? 

Now, in my last estimate of Ukrainian losses, I put Ukrainian losses at least at 60,000 killed. I am not sure I am ready to back away from that: See: The Russo-Ukrainian War – Day 699 | Mystics & Statistics (

Part of the reason for my higher totals is that I don’t think there is a wide disparency between Ukrainian and Russian losses. Yet, Zelenskyy is saying there is an exchange rate of almost 6-to-1. How does that happen? Supposedly Russian has an artillery advantage, Ukraine has a shell shortage (which is part of the reason they are claiming for the loss of Avdiivka), and Russia is firing like three times to ten times the rounds per day as Ukraine is. Now, in a modern conventional war (modern in this case is defined as 1904 and later), the majority of casualties are taken by artillery and other high explosive weapons. If Russia has a significant artillery advantage, then how in the hell are they taking six times the losses of the Ukrainians? This simply defies logic. 

So, either the Ukrainian claims of losses is too low (which is probably the case) or the Ukrainian claims of Russian losses are simply too high (which is also probably the case). I will stick with my previous estimates (repeated liink): The Russo-Ukrainian War – Day 699 | Mystics & Statistics (

Now, the last report from the Ukrainian defense ministry was that Russian losses were over 400,000 killed (409,820 as of 25 February 2024). This number had been batted around without comment by a lot of people on twitter (now known as X – which sucks as an App name). Clearly, Zelenskyy is now repudiating his defense ministry’s own claims by providing a figure that is less than half of that. Are all those people on twitter who have accepted these previous Ukrainian claims lock, stock and barrel going to scale back to the lower figures? I am tempted to call them out by name, but I prefer to find the truth, not get into arguments with people. There has been a lot of really light-weight bullshit casualty figures put out there by people who should know better.

So let us say that Russia does have an artillery advantage and has been firing more shells (see: The West is underestimating Ukraine’s artillery needs – Defense One). That would argue for higher Ukrainian losses than Russian losses. Now, I gather the nature of their artillery shells are not the same. Russia is firing lots of “dumb” munitions. That is certainly the case with the stuff they are getting from North Korea. I gather the percent of rounds that are smart munitions is higher (much higher?) for the Ukrainians than for Russia. So, less firepower but more accuracy. Does that generate a more favorable exchange ratio for Ukraine? Could be. Does in generate a 6-to-1 exchange ratio? Probably not.

Now, for all practical purposes, Ukraine has really not been on the offensive since November 2022. The little attack that they did in early June 2023 was “offensive light” and called off before the casualties got too high. They really have not done anything since then. On the other hand, Russia has had two extended furious offensives, at Bakhmut in early 2023, where they clearly took losses (and threw people away in penal units) and now Avdiivka. Do these two offensives add up to another 150,000 killed above and beyond what Ukraine has lost? I seriously doubt it. I do not rule out that it did run up Russian losses to be higher than Ukrainian losses, which is part of the reason I potentially put Ukrainian losses at only 75% of Russian losses. This a very different than a 6-to-1 exchange ratio.

So, for now I am sticking with my estimates as outlines in the post (repeated link): The Russo-Ukrainian War – Day 699 | Mystics & Statistics ( The count of Russian losses by Mediazona is now 44,654 (as of 15 February 2024). If I double that, then we are looking at potentially 89,308 killed. This is about half of what Zelenskyy is saying. Now, the Ukrainian Book of Remembrance for the Fallen for Ukraine had 22,224 names listed of the end of August 2023, while Mediazone has 30,698 Russian names listed as of 24 August 2023 (this is 72%). If we take 75% of Russia’s casualties, then we end up with 66,981. Is this a valid Ukrainian casualty estimate? If so, that puts Ukrainian losses at over twice what Zelenskyy is saying.

Is it possible in the middle of a war that a politician running one side could claim their losses are half of what they actually suffered while doubling their enemy’s losses? Distinctly possible.

This entry was posted in Eastern Europe, Russia by Christopher A. Lawrence. Bookmark the permalink.

About Christopher A. Lawrence

Christopher A. Lawrence is a professional historian and military analyst. He is the Executive Director and President of The Dupuy Institute, an organization dedicated to scholarly research and objective analysis of historical data related to armed conflict and the resolution of armed conflict. The Dupuy Institute provides independent, historically-based analyses of lessons learned from modern military experience. ... Mr. Lawrence was the program manager for the Ardennes Campaign Simulation Data Base, the Kursk Data Base, the Modern Insurgency Spread Sheets and for a number of other smaller combat data bases. He has participated in casualty estimation studies (including estimates for Bosnia and Iraq) and studies of air campaign modeling, enemy prisoner of war capture rates, medium weight armor, urban warfare, situational awareness, counterinsurgency and other subjects for the U.S. Army, the Defense Department, the Joint Staff and the U.S. Air Force. He has also directed a number of studies related to the military impact of banning antipersonnel mines for the Joint Staff, Los Alamos National Laboratories and the Vietnam Veterans of American Foundation. ... His published works include papers and monographs for the Congressional Office of Technology Assessment and the Vietnam Veterans of American Foundation, in addition to over 40 articles written for limited-distribution newsletters and over 60 analytical reports prepared for the Defense Department. He is the author of Kursk: The Battle of Prokhorovka (Aberdeen Books, Sheridan, CO., 2015), America’s Modern Wars: Understanding Iraq, Afghanistan and Vietnam (Casemate Publishers, Philadelphia & Oxford, 2015), War by Numbers: Understanding Conventional Combat (Potomac Books, Lincoln, NE., 2017) , The Battle of Prokhorovka (Stackpole Books, Guilford, CT., 2019), The Battle for Kyiv (Frontline Books, Yorkshire, UK, 2023), Aces at Kursk (Air World, Yorkshire, UK, 2024), Hunting Falcon: The Story of WWI German Ace Hans-Joachim Buddecke (Air World, Yorkshire, UK, 2024) and The Siege of Mariupol (Frontline Books, Yorkshire, UK, 2024). ... Mr. Lawrence lives in northern Virginia, near Washington, D.C., with his wife and son.

17 thoughts on “So, has Ukraine only lost 31,000 since the start of the war?

  1. Can we infer anything from the vehicle/equipment loss ratios onto the casualty rates? Intuitively it feels like there should be some kind of relationship, and the vehicle and equipment losses seem much better documented?

    • Well, I think the high vehicles losses has led some people to accept the very high estimates of Russian losses. The only report I am aware of that connects vehicle losses to personnel losses in the Russian First Guards Tank Army report of 24 February-15 March 2022. They are clearly losing less than one person (wounded or killed) per tank lost.

      Those reports are here and here

      So yea, I don’t think one can draw a straight line from vehicle losses to personnel losses unless there was more data. With less than one personnel casualty per vehicle destroyed, that does not get you anywhere near the losses that have been reported or even the much higher figures of losses that are claimed.

      For WWII data, Western Front, Trevor Dupuy and HERO did do a correlation tying equipment losses to personnel losses. The Army combat modeling community then reversed those tables (estimating personnel losses based upon equipment losses) and used that for the combat model calculations. In many cases, their combat models only track equipment losses (because often in their hierarchy of model structures they are using a tank vs tank and gun vs tank tactical model as their starting point). Therefore, they needed some way to calculate personnel losses, so they used WWII data generated by HERO and reversed the relationship. There is probably the basis for a long blog post here. There is probably the need to an extended study on this subject, as I don’t think any of this data had been updated since the 1970s.

      • Thanks. Yes, I didn’t expect there to be necessarily a really obvious relationship, just ‘some’ relationship. Am I right in recalling that Allied tank crews had a vehicle to KIA ratio of a little above 1 (and that got closer to 1 after some time of the unit in combat). But perhaps crew survivability, willingness to bail, and the apparent relatively small proportion of tanks killed by large direct fire weapons are causing historically quite low ratios? In any case, I was thinking of them more as a broad ‘intensity’ measure.

        Really enjoyed the latest book BTW.

        • I have heard people claim that Allied tank crews in WWII, have a 1-to-1 wounded-to-killed ratio, but I don’t think that statistic came from our work. At least, I can’t find it in our work. I do wonder where that statistic comes from (assuming it is correct). One will note that that the Russian First Guards Tank Army had a wounded-to-killed of 2.88-to-1, but this clearly includes people who are outside of tanks (lots of people in armor units get wounded outside of their vehicles).

          Glad you like the book (I assume you are talking the Kyiv book).

  2. Chris, I wonder how much of Russian artillery fire was aimed at civilian targets rather than military targets (even though I’m guessing that most targeting at civilian targets was with bombs/rockets/missiles and drones). Anyway, if a politician (or a military officer) gives a number for killed (not lost) without defining terms and data source then skepticism seems reasonable. Killed on the battlefield (not counting those later dying from wounds received on the battlefield)? Excluding deaths from disease and accidents? Probably better to wait for a more detailed statement rather than give much credence to a figure offered during a speech.

  3. C.A.L: “First, who makes up those 31,000 killed? If Zelenskyy has thrown out a figure for the first time since the summer of 2022, I am guessing it is based upon something…”

    -I was a little skeptical, too.

    You could also ask, “31,000 killed, as of when?”

    John: “Can we infer anything from the vehicle/equipment loss ratios onto the casualty rates? Intuitively it feels like there should be some kind of relationship, and the vehicle and equipment losses seem much better documented?”

    -Attrition gave the stat’s as 1% of main battle tank losses for every 1% human battle casualties if the force had 1 MBT per 1,000 troops, increasing by 1% for every additional MBT per 1,000 troops (e.g. 3% human losses @ 3 MBTs per 1,000 troops) up to a max’ of around 6% at 6 MBTs per 1,000. I mentioned this a while ago on this blog, but I can’t find it.

    I think the ratio for the Russians and Donbas allies has averaged something like 2 MBTs per 1,000? If so, the ratio would be half the rate of MBT losses.


    James D. Glick
    PO1, USNR (ret.)
    Clarksville, TN
    “Hey! What happened to all my silverware?” -Lady Selkirk, 1778.

  4. A link to the graphs for the ratio of human vs. tank losses in a large force in Attrition:


    A link to the chart for the ratio between tank losses (inc. light tanks) and crew losses:

    …but I wonder if the ratio might be worse for Soviet/Russian tanks vs. M4 Shermans and Cromwells? Not that the later were particularly safe by the standards of an M1A2 Abrams…

    James D. Glick
    PO1, USNR (ret.)
    Clarksville, TN
    “Hey! What happened to all my silverware?” -Lady Selkirk, 1778.

  5. Doh! I don’t know what sort of math was going on in my head: There were (maybe) 190,000 Russian troops and (maybe) 44,000 Donbas troops in FEB 2022. 1,400 MBTs would give a ratio of 6 MBTs per 1,000 troops, so tank losses would be 6 times human losses, or put the other way, human losses would be one-sixth of MBT losses.

    BTW, I’m only counting MBTs, although I think Dupuy was counting light tanks as well. If you count them (e.g. BRM recon vehicles), the ratio would be even higher, past the plateau of diminishing returns. It’s also possible that the ratio may have dropped in those cases when the Russians have been dug in.

  6. Another brain fart on my part. The original topic was Ukrainian losses. I think the number of MBTs per 1,000 was relatively low, but still over 1 per 1,000. In the recent Summer Offensive, it was probably over 6 per 1,000, at least at first.

    James D. Glick
    PO1, USNR (ret.)
    Clarksville, TN
    “Hey! What happened to all my silverware?” -Lady Selkirk, 1778.

  7. In response to the post commenting on Zelensky’s 31,000 KIA. In that article you suggest 60,000 Ukrainian KIA and 89,000 Russian KIA. How do you account for the well documented asymmetry in artillery and airpower?

    Personally, I’ve seen the performance of too many US and NATO trained ground forces that I wouldn’t be surprised if Ukraine was taking three times your number of KIA. Furthermore, with a casualty to dead ratio of four or five to one. The resulting number of casualties would explain the shortage of military age males and the 40% strength of many battalions on the front lines in the Donbas reported by the WaPo.

  8. UALosses project documented 44,712 Ukrainian service members by name who have died in the war between 24 February 2022 and 10 March 2024. Another similar project, the Book of Memory, recorded 24,500 all-causes deaths by name from 24 February 2022 to 14 November 2023, with the statement that the real number was likely around 35,000 (i.e, those who could be identified are “not less” than 70% of the actual total), plus 15,000 missing, 3,400 captured, and about 100,000 wounded. (Yaroslav Tychenko and Herman Shapovalenko in the journal “Tyzhden”)

    If UALosses’ more recent estimate likewise accounts for only 70% of Ukrainian military deaths, it suggest, as you wrote, that the real total is about 65,000, with proportionate numbers of wounded and missing.

Leave a Reply

Your email address will not be published. Required fields are marked *