How many research stations are there in Antarctica? The list this site uses carries 114 facilities, run by 31 nations. That is the number to quote, and the reason no two sources agree on it is sitting inside the number: of those 114, only 83 are classified as stations. The other 31 are camps, refuges, airfield camps, depots and laboratories, and whether you count them is a choice rather than a fact.
Three more cuts change the answer again. 103 of the 114 are in use and 11 are temporarily closed. Only 44 of them stay open through the winter. And nobody publishes how many people are inside any of them in July. So the honest answer is a set of four numbers rather than one, and this post is the set.
The source, and why it is the one to use
The register is the COMNAP Antarctic Facilities Master, maintained by the Council of Managers of National Antarctic Programs and republished as open data by the Polar Geospatial Center. COMNAP is the body the national programmes themselves belong to, so the list is a self declaration by the operators rather than an outside survey. It is free, it needs no key, and it carries a position, an operator, a facility type, a season and an establishment year for each entry. Every one of the 114 has its own page on this site, and they are all on the live map.
114, split four ways
| facility type | count |
|---|---|
| Station | 83 |
| Camp | 10 |
| Refuge | 8 |
| Airfield camp | 7 |
| Depot | 3 |
| Laboratory | 3 |
A depot is a fuel cache. A refuge is a hut you can survive in. An airfield camp can be a graded strip of blue ice with a tent beside it. None of those is what a reader picturing Antarctic research has in mind, and all of them turn up in station counts published elsewhere, which is most of why the figure floats between about 70 and about 120 depending on who you ask. Count only the buildings COMNAP itself calls stations and you get 83.
The split that actually matters is the season
44 facilities are year-round and 70 are seasonal. Put the 11 temporarily closed ones aside and 59 of the 103 places in use shut when the light goes. That is well over half the human presence on the continent packing up every autumn, and it does not thin the map evenly, it empties whole sectors. We measured what that does to the geometry: the furthest you can stand from anybody drops from 1,815 km in winter to 924 km in summer, and the median patch of mapped ice goes from 515 km to 300 km from the nearest facility. The winter map is not a sparser version of the summer map. It is a different continent.
Who runs them
| operator | facilities | year-round |
|---|---|---|
| Chile | 14 | 5 |
| Argentina | 13 | 7 |
| Russia | 11 | 5 |
| Australia | 7 | 4 |
| United States | 7 | 3 |
| China | 6 | 3 |
| Italy | 6 | 1 |
| Brazil | 5 | 1 |
| Japan | 5 | 1 |
| United Kingdom | 5 | 1 |
Chile has the most facilities and Argentina has the most that stay open all year. Both counts are concentrated in one place: the Antarctic Peninsula and the islands north of it, which is the only part of the continent within reach of a short flight or a working port. The United States runs 7 facilities and one of them, Amundsen-Scott at the geographic pole, is the nearest thing to a person for 1.52 million square kilometres of winter ice, which is more ice than any other facility on the list can claim.
The ranking by count and the ranking by reach are different rankings, and the second one is the interesting one. Ten facilities are between them the nearest thing to 72.6% of the mapped ice in winter.
The oldest and the newest
97 of the 114 carry a founding year. The oldest is Orcadas, on Laurie Island in the South Orkneys, established in 1904 and staffed by Argentina through every winter since. It predates the Antarctic Treaty by 55 years. The newest is Qinling, a Chinese year-round station at 74.93 south, 163.71 east on the Ross Sea coast, dated 2024. The next three oldest after Orcadas are all from 1947, and the busiest founding years on the whole list are 1957 with five and 1956 with four, which is a fair sketch of when the modern continent was staked out.
The number nobody has
The most common follow up question is how many people are down there, and the answer is that no defensible figure exists. COMNAP records a peak population for 93 of the 114 facilities and a winter population for none of them. Peak population is a capacity figure reported by the operator, not a census, and it is undated. Adding 93 capacity figures together and calling the total a headcount would be inventing a number, so this site does not print one. Estimates in the range of a thousand people in winter and four thousand in summer circulate widely, and every one of them traces back to the same undated capacity figures.
Saying so is more useful than guessing. The count of open buildings is measured. The count of people inside them is not.
The short version
| question | answer |
|---|---|
| facilities on the COMNAP list | 114 |
| classified as stations | 83 |
| in use | 103 |
| temporarily closed | 11 |
| open through the winter | 44 |
| nations operating them | 31 |
| with a published winter population | 0 |
If you need one figure for a headline, use 114 facilities run by 31 nations, and say facilities rather than stations. If the sentence is about people or about science actually being done in the dark, use 44. The gap between those two numbers is the single most misunderstood thing about the continent, and it is bigger than the gap between any two published station counts.