Ask an Analyst: Putting the pieces together on military troop data
Go behind the scenes with our team as we find and make sense of the numbers.
Every time there is a new development in foreign affairs, be it a military operation in Venezuela or the US response on Russia’s invasion of Ukraine, we always get the same question: How many US troops are in or near that country? You would think this is a pretty straightforward question to answer, considering that “How many” questions are usually some of the easiest for government data to answer. But not in this case.
Meet the players
Okay, let's get into it. We use two Defense Manpower Data Center (DMDC) datasets to get the totals for troop data.
The Active Duty Military Personnel by Service monthly tables. These tables give us the total number of active-duty members under the Department of Defense by branch.
There is a lot of good information here! For right now, we only really need the “Grand total” row.
“But wait!” you ask. “What about the Coast Guard?” The Coast Guard actually operates under the Department of Homeland Security, so it is not included in this total. (Fun fact: In times of war or at the president’s order, the Coast Guard serves as a part of the Navy and the Defense Department.)
So, we’re close to getting a complete number of active-duty troops! We’re just missing the number of Coast Guard active-duty troops. Enter dataset 2.
The Military and Civilian Personnel by Service/Agency by State/Country quarterly tables, or as we call it, the motherlode. This dataset provides unclassified locations of active-duty troops in the US, its territories, and abroad. We have the unclassified number for troops in 173 countries, all 50 states, Washington, DC, and six US territories.
However, we can’t get geographic data on each service member’s location. My guess is that it’s an issue of national security. The US government wouldn’t want foreign adversaries to have specific data on troops’ locations. Also, the dataset may not include troops on temporary duty or those sent in support of contingency operations. For example, the troops that were in Venezuela in early January would most likely not show up in the quarterly troop location report.
So, what we gain in location, we lose in completeness: this dataset is not as comprehensive as the Active Duty Military Personnel by Service table. But it does provide an idea of Coast Guard levels. Interestingly, it does have complete totals for military reservists.
So, if you’ve wondered why our maps of active-duty troop levels (which add up to 1,326,688 in September 2025 from the location report) don’t match up with the total number of active-duty troops (1,339,703 from the active duty report and Coast Guard total), now you know!
Putting 2 and 2 together:
To calculate total active-duty troops by military branch, we use the active-duty report and add in the Coast Guard value from the location report. Separately, the complete count of reserve troops is in the location report, so we don’t need to grab that from somewhere else (although it does exist in its own report as well).
TA-DA!
Troops in the armed forces, by type and branch (September 2025)
Want to take it a step further? Download both of these reports for each fiscal year (yes, they are all provided in single Excel files and not in a time series *eye roll*), and then you can see how the totals change over time! To be honest, I didn’t do all this myself, luckily our data engineers were able to pull that data together a lot faster than I could.
Troops in the armed forces, by type and fiscal year
Can we see changes in the data after international conflicts?
Let’s take a look at a real-world example. In February 2022, Russia launched a full-scale invasion of Ukraine. Readers asked “How many troops are in/or around Ukraine?” question more times than I can count, and unfortunately, we couldn’t really answer the question. (Fast forward to now, and we’re getting the same questions about Venezuela and Cuba). At the time, the most recent report we had (from December 2021) pointed out that there were 41 troops in Ukraine. This seemed like an incomplete story, though. After the DMDC released some more reports, we were able to make a time comparison.
The time comparison I’m making is between the report from December 2021 and the one from September 2022. Ideally, I would want to compare using the December 2022 report, but the Army did not provide personnel data in that release. Under normal circumstances, I would use the same reporting period to compare. Troop counts might change throughout the year due to rotations and reporting cycles, and mismatched timing can make normal fluctuations look like real change. However, in this case, I decided to use complete data from a different reporting period rather than compare incomplete data.
We went from having 41 troops in Ukraine to seven, which struck me as odd. Was there another angle we could look at the data through?
Turns out, yes.
A June 2022 news release from the Biden administration’s Defense Department explained that the US’ response involved cooperation with NATO countries in Europe, like Germany and Poland. The release states, “The DoD deployed or extended over 20,000 additional forces to Europe in response to the Ukraine crisis, adding additional air, land, maritime, cyber, and space capabilities, bringing our current total to more than 100,000 service members across Europe.”
Like we learned previously, our main data source does not provide data on troops deployed temporarily (2022 was the first year of the Russia-Ukraine war and these deployments might have been thought to have been temporary in nature.) or in support of contingency operations. So, I think it's pretty safe to say that data from that time will most likely not reflect those additional 20,000 troops, but let's take a look anyway.
From December 2021 to September 2022, the total number of US troops in NATO countries in Europe increased by 481. Not quite the 20,000 we’re looking for. However, the number of troops increased the most in Germany and the United Kingdom by 223 and 221, respectively. Out of the 30 European NATO countries, troops increased in 20 of them.
I put together the quarterly data from Q4 2017 through Q3 2025 into a map (and excluded Q4 2022 through Q2 2023 because of the incomplete reporting) to see if there were any visual takeaways. Conclusion: I don’t see much meaningful change. (Another side note: We at USAFacts will follow the data wherever it leads us, and we’ll experiment with different ways of thinking about it. In this case, there wasn’t an interesting takeaway from the data story, but I had to try it nonetheless!) The context is there, though: the US has a large concentration of troops in Germany, which is pretty consistent over time.
Troops levels in European NATO countries, Q4 2017 through Q3 2025
Therefore, official statements describing force increases may reflect a mix of temporary movements and assignments, which are not always visible in declassified troop location data.
So… is this data actually useful?
Short answer: yes, just not in the way we might want it to be.
The troop location data isn’t meant to tell us what’s happening right now during an active conflict. And frankly, it makes sense. Publishing real-time or highly specific troop movements would be a national security nightmare.
Where this data really shines is in showing the big picture. Because the reports are consistently released and (mostly) standardized, they’re best used to understand where the military maintains a standing presence and how that footprint changes over time.
Think of it as a baseline for what “normal” looks like. When something major happens, like a war or a surge in activity, you can then pair this data with official announcements or government statements to better understand what’s changing and why.
The question “How many troops are in a country?” sounds straightforward, but as we’ve seen, it rarely is. While this data doesn’t provide a real-time answer, it does give us the context in which those questions sit, and that’s a critical part of the picture.
