Everest Has Killed 330 People — Here's the Denominator Nobody Mentions

330 Deaths. Out of How Many?

Every May, when the weather windows open on the world’s highest mountain, the same numbers resurface. More than 330 people have died on Everest since the first British expeditions in the 1920s. Around 200 of the bodies are still up there, frozen into the route, some of them used as landmarks. The image of a queue of climbers waiting their turn beside a corpse has become the mountain’s signature photograph.

The body count is real. The Himalayan Database, which has tracked every expedition since 1905, puts the total a little above 330 as of the 2023 season. What the number lacks is a denominator. Three hundred and thirty deaths across a century, or across twelve thousand summits, or across the twenty-five thousand or more people who have climbed above base camp? Each produces a different picture, and the headline never says which one it means.

The 1% Figure and Where It Comes From

The second famous number is a rate: one in a hundred people who attempt Everest die. It comes from actual research rather than from a press release, which makes it better than most of the statistics on this site, but it still hides more than it shows.

A 2008 study in the BMJ went through every death on the mountain from 1921 to 2006 and matched them against everyone who had climbed above base camp. The result was a mortality rate of roughly 1.3%. A 2020 follow-up in PLOS ONE by Raymond Huey and colleagues covered 1990 to 2019 and found that while summit success had doubled (from about one in three climbers in the 1990s to two in three after 2006), the death rate had stayed close to 1%. Better forecasting, fixed ropes and bottled oxygen got far more people to the top without making the top any safer to stand on.

So one in a hundred is a fair average. The question is who it is an average of.

Two Different Mountains

The BMJ study found that Everest kills its two populations in different places and for different reasons. Paying climbers, the “members” in Himalayan Database terminology, die mostly above 8,000 metres, and mostly on the way down. The pattern of symptoms in those deaths (confusion, loss of coordination, collapse at fixed points) points to high-altitude cerebral oedema and exhaustion rather than to falls or weather. They reached the summit, or turned around close to it, and did not have enough left for the descent.

Sherpas and other high-altitude workers die lower on the mountain, in the Khumbu Icefall, where the glacier shifts and collapses without warning. They cross it far more often than clients do, ferrying loads and fixing the route. In 2014 an ice avalanche in the Icefall killed sixteen Nepali workers in a single morning. The following year, the earthquake that hit Nepal sent an avalanche off Pumori into Base Camp and killed at least eighteen people, the deadliest day in the mountain’s history.

It would be tempting to assume the workers’ deaths inflate the clients’ rate. They do not. The BMJ study counted 8,030 climbers and 6,108 Sherpas and found the climbers died at 1.6% and the Sherpas at 1.1%; the 1.3% headline sits between them. On the descent from the summit the gap widens to 2.7% for climbers against 0.4% for Sherpas. Huey’s 2020 rate of about 1% covers members only, no Sherpas at all. The paying client is, if anything, the more exposed of the two populations. What the average hides is where the risk lives: for a client it is the last 800 metres, for a worker it is a section the client passes through four times. The average also blends eras. A 1930s expedition in tweed and leather with no fixed ropes is not the same activity as a 2024 guided climb with oxygen from Camp 3 and a weather forecast on a satellite phone, yet both feed the same century-long rate.

The Chain of Probabilities

The calculator below takes the 1%-per-expedition figure and spreads it across the days a climber spends above base camp. A standard commercial expedition runs about 40 days on the mountain, including acclimatisation rotations up to Camp 3 and back. One percent over 40 days works out to a daily death rate of roughly one in 4,000.

Spreading the risk evenly is a simplification. Almost nobody dies on a rest day at Camp 2. The real risk sits in two places: every crossing of the Icefall, and the roughly 36 hours of the summit push. But the daily rate does capture the thing the headline misses, which is that the risk scales with exposure. A climber who does two acclimatisation rotations instead of one crosses the Icefall two extra times. A season with a narrow weather window forces everybody through the same 48 hours, which is how the 2019 queue happened, and how 11 people died that year.

Your Actual Risk

The two sliders are how many expeditions you join this year and how many days each one keeps you above base camp. For the standard one-expedition, 40-day profile, the result is 1-in-100, matching the research. A fast, well-acclimatised climber who spends 25 days on the mountain comes out closer to 1-in-160. Somebody guiding two expeditions in a season, or spending 70 days on a slow ascent, moves past 1-in-60.

For scale, a climber who does one expedition in a year spends that year at about eighty times the annual risk of dying in a car crash in the United States (1-in-100 against roughly 1-in-8,000). It is a risk profile that almost no other activity available to a paying customer comes close to. And it is the risk for a single climb, not a lifetime.

For everyone who does not climb Everest, which is roughly eight billion people, the risk is zero. The body count tells them nothing.

What the Body Count Gets Wrong

The 330 figure keeps circulating because it sounds like a warning. It functions as the opposite. Divided across a century and a famous mountain, it reads as history, a list of names on a memorial rather than a statement about the person booking a permit next spring.

The rate is the warning. One in a hundred, per climb, in the best-equipped era the mountain has ever seen. The people who should hear that are not the general public, who will never go, but the growing number of clients who see the summit-success rate climb toward two thirds and assume the danger has fallen with it. It has not. More people are reaching the top, and the same fraction are not coming back.

Sources

• The Himalayan Database (Elizabeth Hawley, Richard Salisbury et al.). Expedition records for the Nepal Himalaya since 1905. Cumulative Everest deaths and summits through the 2023 season.
• Firth, P.G., Zheng, H., Windsor, J.S. et al. (2008). Mortality on Mount Everest, 1921-2006: descriptive study. BMJ, 337, a2654. 212 deaths among 14,138 mountaineers above base camp (8,030 climbers, 6,108 Sherpas); mortality 1.3% overall, 1.6% for climbers, 1.1% for Sherpas; 2.7% vs 0.4% on descent from the summit via standard routes. Members died mainly above 8,000 m during descent; Sherpas mainly in the Khumbu Icefall.
• Huey, R.B., Carroll, C., Salisbury, R., Wang, J.L. (2020). Mountaineers on Mount Everest: Effects of age, sex, experience, and crowding on rates of success and death. PLOS ONE, 15(8), e0236919. Summit success rose from about one third to two thirds between 1990-2005 and 2006-2019; death rate stayed near 1%.
• Nepal Department of Tourism. 2023 permit count (478, a record) and 2023 season deaths (18).
• Reporting on the 18 April 2014 Khumbu Icefall avalanche (16 killed) and the 25 April 2015 earthquake avalanche at Base Camp.