What the climbing accident data shows, and who collects it.

Who collects the data

North America: the American Alpine Club and the Alpine Club of Canada

The American Alpine Club (AAC) says it has published Accidents in North American Climbing every year since 1948, analysing the year's most significant and instructive accidents. Its online archive holds more than 5,000 reports, free to read. Reports come in through an online form, from climbers themselves or from rescuers, rangers and others who know the incident, and they may be anonymous. The title has changed over time: editions into the 2010s were called Accidents in North American Mountaineering, and the 2024 and 2025 editions carry the climbing title. The AAC describes the 2025 edition, published in September 2025, as covering rock, alpine and ski mountaineering accidents in the United States, Canada and Mexico, with more than 100 incidents analysed. Pete Takeda edits it.

The Alpine Club of Canada (ACC) wrote in 2019 that it has contributed the Canadian material since 1977. The AAC's own 1966 issue, though, already called itself the sixth produced in conjunction with the ACC. The two may be describing different kinds of involvement; the sources do not say.

United Kingdom: the BMC and mountain rescue

The British Mountaineering Council (BMC) opened its Incident and Near-miss Reporting System in April 2019. Reports are voluntary, moderated and published to a free database. In July 2019 the BMC reported more than 100 reports in the first three months. A BMC analysis of the trial year's themes, which the BMC said would follow April 2020, could not be found.

Rescue services count call-outs rather than climbing accidents. A BMC article citing Mountain Rescue England and Wales (MREW) gives 3,462 calls and 2,775 team deployments in 2023. A North East Wales team, reporting MREW figures on 31 March 2026, gives 3,968 call-outs in 2025 and 3,784 in 2024. A cycling magazine gives 3,842 call-outs for 2024 instead; which 2024 figure is right is unconfirmed. No MREW figure isolating climbing was found.

Scottish Mountain Rescue's 2024 statistics report (Jessica Steinemann and Andy Morgan, April 2025) does break activities down. Of 636 incidents in 2024, 335 were mountaineering in its broad sense. Of those, 11 were rock climbing, 10 scrambling and none snow or ice climbing; most were hillwalking. The report itself is not consistent in two places: winter hillwalking appears as 24 percent in a table and 27 percent in a chart, and days with at least one call-out appear as both 287 and 278.

The Alps: Switzerland, Austria and Germany

The SAC compiles its figures from Rega, Alpine Rettung Schweiz, the Valais rescue organisation and other bodies; the figures here come from press reports, because the SAC's own statistics files could not be opened. The OeKAS release has its own inconsistencies: one table header gives 266 deaths for 2024 against the 309 total, and the 10-year mean for people involved appears as both 12,033 and 12,034. The German Alpine Club (DAV) publishes member accident statistics, but its site returned errors and its figures could not be confirmed; its gym statistics with the gym association KLEVER are on the gym safety page.

How each body ranks the causes

The bodies use different categories and study different people, so their rankings cannot be merged. Each is given here as published.

A search summary of a DAV article on the 2025 Swiss figures says falls were the most common cause of rescue across mountain sports, but that being stuck was the most common reason for climbers and high-alpine parties; the page itself returned an error, so this is unconfirmed.

The long-run studies

Caroom (2020). Eliot Caroom's analysis for Rock and Ice (issue 265, September 2020; republished by Climbing in February 2022) read 2,770 ANAC narratives from 1990, about 56 percent of ANAC accidents in that time. The same article also describes its base as 2,724 incidents; the two counts are not reconciled. Trad climbing outside alpine terrain made up 63 percent of roped accidents and sport plus top-rope 20 percent. Caroom says this does not show trad is three times as dangerous, because the sample is not random and nobody knows how many days are climbed in each style. He tagged several factors per incident, unlike ANAC's single immediate cause, so his percentages do not line up with ANAC's tables. He also notes that overuse injuries are not covered and that gym and bouldering accidents are rarely reported to ANAC.

Curran-Sills and Karahalios (2015). Their study of ACC search and rescue records from 1970 to mid-2005 covered 1,088 incidents and 1,377 casualties, of whom 433 (31.4 percent) died. Mountaineering, skiing, ski mountaineering and snowboarding made up almost two thirds of deaths, and British Columbia and Alberta 91.6 percent of incidents.

Mort and Godden (2010). Across UK mountain rescues from 2002 to 2006, 6,814 incidents involved 7,995 people, including 550 fatal incidents. Hillwalking accounted for 75 percent of rescues.

Rugg and colleagues (2020). From the Austrian National Registry of Mountain Accidents, they counted 2,992 rock climbing incidents from 2005 to 2018 and 140 deaths, a case fatality rate of 4.7 percent. The injuries page has their injury findings.

Why rates are hard to give

An accident rate needs a count of people climbing, and that count is itself disputed. In the 2003 report, the ANAC editors quoted an Outdoor Industry Association estimate of six million people climbing on natural rock, then gave their own estimate of 250,000 to 300,000 enthusiasts, noting how much this changes any rate. The 1955 report already said many accidents probably went unreported, and the 2003 editors said the same of Canada. For participation figures, see how many people climb. For the patterns in roped climbing specifically, see roped climbing incidents.