K2 Death Rate: How Many Climbers Have Died on K2? (2026)

The K2 death rate is about 9.5% when measured as deaths per successful summit: 92 recorded deaths against 964 recorded summits through December 31, 2025, or 92 ÷ 964 = 9.54%. That is a death to summit ratio, not the chance that a climber dies on any one attempt. The famous “one in four” figure (25%) was a fair summary of older tallies, but it described one death for every four summits in an era when very few people had reached the top.

The rest of this guide explains where each number comes from, why sources disagree, and what the statistics can and cannot tell you. Every percentage below is paired with its numerator, denominator, and cutoff date.

What Is the K2 Death Rate?

K2 has no single official death rate. Different sources calculate different ratios, and each answers a slightly different question.

The most common version divides recorded deaths by recorded successful summits. Using the 1930 to 2025 compilation this guide relies on, built from 8000ers.com, ExplorersWeb, and Alan Arnette’s season reporting, that is 92 ÷ 964 = 9.54%. Put another way, there has been about one recorded death for every 10.5 successful summits.

This is a death to summit ratio, not the percentage chance that a climber dies on a particular attempt. Climbers who turn back are not included in the denominator.

Other measures exist, and mixing them up causes most of the confusion online:

Metric	Calculation	What it tells you	Main limitation
Death to summit ratio	Deaths ÷ successful summits	How the historical death toll compares with summit volume	Not a personal probability; climbers who turned back are missing
Deaths per attempt	Deaths ÷ all serious attempts	Closest historical measure to per attempt risk	No reliable all history attempt count exists
Summiter fatality rate	Climbers who reached the summit and died during the descent ÷ all climbers who reached the summit	Descent risk after reaching the top	Needs individual level records
Season ratio	One season’s deaths ÷ that season’s summits	A single year’s snapshot	Swings wildly; undefined in a zero summit year
Absolute death toll	Total recorded deaths	The human cost	Ignores how many people climbed

When you see “K2 fatality rate,” “K2 mortality rate,” or “death rate on K2” in an article, check which row of this table the author is using. Most are quoting the first one.

Why Is K2’s Famous 25% Death Rate Outdated?

The “one in four” figure came from a small denominator, meaning a small number of summits. Wikipedia notes that before 2021, roughly one person had died on K2 for every four who summited, and a March 2012 tally of 81 deaths against 306 ascents works out to 26.5%. Within the 1930 to 2025 compilation used here, the period through 2000 shows about 47 deaths against about 164 summits, roughly 29%. These are different datasets that count deaths and summits differently, so they should not be read as one continuous series. What they share is the pattern: for decades few people summited K2, so a handful of disasters made the ratio look extremely high.

Then the number of summits grew. Commercial expeditions, better weather forecasting, fixed ropes, and wider use of supplemental oxygen are often credited with helping more climbers reach the top, especially since 2018. Around 800 of the 964 recorded summits came after 2000, and 2022 alone produced roughly 200. Over that same later period, the compilation counts about 45 deaths, or about 5.6 per 100 summits.

That is why the ratio moved from about one death for every four summits to about one for every ten: the count of summits grew much faster than the count of deaths. Some in between figures you will see, such as Wikipedia’s estimate of 96 deaths against about 800 summits as of August 2023 (about 12%), reflect different cutoff dates and counting rules and not a contradiction.

This does not mean K2 became a safe mountain. The Bottleneck is still there, and so are the storms, the rockfall, and the long technical descent. The ratio fell as summit numbers rose. Better logistics and support may also have helped, but the data cannot show how much.

How Many People Have Died on K2?

The compilation used in this guide records 92 deaths through December 31, 2025, and that is the figure used throughout this article. Wikipedia’s list of deaths on eight thousanders gave an estimate of 96 deaths against about 800 summits as of August 2023. Sources can differ in how they count deaths, so this guide sticks to one dataset and always states its cutoff date.

Historical lists can differ on several points:

• Whether approach deaths (people who died walking to Base Camp) are counted
• How high altitude workers are classified
• Whether missing climbers are recorded as deaths
• Whether repeat summits count separately
• Later corrections to disputed summit claims

The practical advice is to trust any number that states its cutoff date and source. A page that gives a total with no date and no source is giving you less than it appears to.

The 2026 Season, Kept Separate

The summer 2026 season is not part of the calculation above. Reporting from Dawn (August 16) and Alan Arnette’s season wrap (August 22) indicated that no one summited K2 in summer 2026. Arnette estimated roughly 50 climbers attempted the mountain (his estimate, not an official permit count), and cited high winds, fresh snow, and rockfall. Reports also said a July 30 avalanche on neighboring Broad Peak killed climbers, reportedly ten, and that the remaining K2 team abandoned its summit push to help with the response. Broad Peak deaths are not K2 deaths, and a season with zero summits cannot produce a meaningful death to summit ratio.

K2 Summits and Deaths by Year

A year by year table shows what the cumulative percentage hides: individual seasons can vary considerably. One bad day can dominate a small season, while a large summit year can make a serious accident look mild.

Year	Summits	Deaths	Deaths per 100 summits	What stands out
2018	62	2	3.2	Record summit year at the time
2019	29	0	0	No deaths recorded
2020	0	0	Not applicable	Season closed because of the COVID pandemic
2021	70	6	8.6	First winter ascent; deaths in winter and summer
2022	200	3	1.5	Record summit year
2023	112	1	0.9	Large season, one death
2024	50	3	6.0	Fewer summits pushes the ratio up
2025 (mid August snapshot)	41	2	4.9	Avalanche and rockfall deaths; partial season

Source: year by year compilation drawing on 8000ers.com, ExplorersWeb, and Alan Arnette data. The 2025 row reflects the source’s mid August 2025 update, not the full year, so late season summits or deaths may not be captured. The headline totals of 92 deaths and 964 summits run through December 31, 2025, and the yearly rows are shown as the compiler lists them, so they may not sum exactly to those totals. The 2026 season is excluded, as explained above.

Taken together, the eight years from 2018 to 2025 in this table (with 2025 covering only the period through mid August) add up to 17 deaths against 564 summits, or about 3.0 per 100 summits. That is far below the historical figure, and it is a good example of how a larger summit count lowers the ratio.

Two lessons come out of this table. First, the low ratios in 2022 and 2023 are partly a crowd effect, since more summits mean a smaller ratio even when the number of deaths is similar. Second, a single bad season like 2008 (18 summits, 11 deaths) or 1986 (27 summits, 13 deaths) can push one year above 40 deaths per 100 summits.

Does K2 Have a Survival Rate?

Not one you can calculate from these figures. If a dataset gives 9.54 deaths per 100 summits, subtracting from 100 to get a “90.46% survival rate” is a mistake.

The reason is the denominator. The ratio counts successful summits, but it leaves out the many climbers who turned around, stopped at a high camp, or never left Base Camp and survived. Those people were exposed to the mountain and are absent from the count. The ratio also compares deaths from all causes and all stages of the climb with a count of people who reached the top, so the two numbers do not describe the same group.

A true per attempt risk would need a consistent definition of an “attempt.” Does a climber who acclimatizes to Camp 3 and goes home count? What about a high altitude worker fixing ropes? Historical records do not answer these questions consistently. So the answer to “what is my chance of dying on K2?” is that no reliable single number exists, and anyone who gives one is choosing a definition without telling you.

Why Is K2 So Dangerous?

K2 is dangerous because several serious hazards stack on top of each other, on a route with little margin for error. No single feature explains the mountain’s reputation. The combination does, and each hazard below shows up in the record.

Extreme altitude

K2 rises to 8,611 meters, above the 8,000 meter line that mountaineers call the death zone. There, the body cannot fully adjust to the thin air, and low oxygen, dehydration, and cold slow movement and cloud judgment. Altitude illness has contributed to deaths on the mountain across many decades, from Dudley Wolfe in 1939 to suspected or confirmed cases in 2022 and 2024.

Technical terrain

The standard Abruzzi Spur route involves steep rock, ice, and mixed climbing much of the way up, with little flat ground for rest. This differs from the standard Everest routes, which rely on more gradual terrain and a larger support system. On K2, a slip that is merely embarrassing at lower altitude can be fatal, and falls appear repeatedly in the death lists of recent years.

The Bottleneck and hanging seracs

At roughly 8,200 meters, climbers enter a steep, narrow gully known as the Bottleneck. Directly above it hangs a massive serac, a tower of glacial ice that can release with little warning. Climbers on the standard route generally pass beneath it, and there is no widely used alternative.

Teams can reduce exposure by moving through early and quickly, but they cannot fully remove the hazard. Fixed ropes also run through this section, so damage to the ropes can strand climbers above it. Damaged ropes were a central factor in 2008, the deadliest single event in K2’s history.

Avalanches and rockfall

Fresh snow, wind loading, and steep slopes create avalanche risk that can close a route entirely. Rockfall has been a recurring problem on the lower and middle mountain in recent seasons. Of the two deaths recorded in the source’s mid August 2025 update, one was an avalanche (July 18) and one was rockfall (August 12).

Extreme weather

K2 sits in the Karakoram, far to the northwest of Everest, and is exposed to sudden, violent storms. Safe summit windows can be as short as a day or two. When a forecast window is missed by a few hours, the consequences can be severe. The 1986 and 1995 disasters both involved storms, and season reporting on 2026 described weeks of wind and snow that left no usable summit window at all.

The descent

Many climbers assume the summit is the finish line, but the hardest decisions often come afterward. Exhausted and short of oxygen, climbers must go back down the same technical terrain, often in fading light.

The best known analysis of this problem was published in the American Alpine Journal by Raymond Huey and Xavier Eguskitza. Looking at K2 summiters from 1978 to 1997, it found:

• About 1 in 7 (13.4%) died during the descent
• 22 of 117 climbers without supplemental oxygen (18.8%) died on the way down, compared with none of 47 who used it
• For comparison, the same study found 1 in 29 (3.4%) of Everest summiters died on descent (1978 to 1999)

Two cautions apply. This is a summiter descent fatality rate from one older era, before commercial expeditions, and it measures a different thing than the 9.54% death to summit ratio, so the two should not be compared. It also cannot tell us how much of the difference was caused by oxygen and how much by the kind of climber who chooses to go without it. I could not find a consistent, verified figure for what share of all K2 deaths happen on the way down, so this guide does not offer one.

Limited rescue

Above the lower slopes, climbers must largely rescue themselves. Helicopters generally cannot operate at the altitudes of the high camps and the Bottleneck, and reaching an injured climber means a long, risky climb by other climbers who are also exhausted. That can turn injuries that would be survivable elsewhere into life threatening emergencies.

The Deadliest K2 Disasters and Major Deaths

A handful of events explain much of K2’s fatality history. Each looks larger against the older, smaller summit counts.

Year	Deaths	What happened
1939	4	American expedition; Dudley Wolfe died, and three Sherpa climbers disappeared
1986	13	The “Black Summer”: avalanches, falls, altitude illness, and a severe storm
1995	7	Storm on the upper mountain; Alison Hargreaves was among the dead
2008	11	Serac collapse in the Bottleneck severed fixed ropes
2021	6	First winter ascent year; multiple winter and summer fatalities

The 1986 disaster

Thirteen climbers died on K2 in the summer of 1986, in several separate incidents involving members of different expeditions on the mountain at the same time. Bad weather, exhaustion, and thinly stocked high camps added to the danger. It remains the deadliest single season on K2, and in the compilation used here it produced 13 deaths against 27 summits.

The 1995 disaster

A violent storm on August 13, 1995 caught climbers high on the mountain, and seven died. British alpinist Alison Hargreaves, who had reached the summit without supplemental oxygen, died on the descent, only months after climbing Everest without oxygen. The event showed that a summit is only half the journey.

The 2008 disaster

Over August 1 and 2, 2008, eleven climbers from several countries died in K2’s deadliest single event. A serac collapse above the Bottleneck destroyed the fixed ropes that climbers needed to descend. With darkness falling and temperatures dropping, several were stranded above the Bottleneck, and further ice falls followed during the night.

It is one of the clearest examples of how one hazard, delays on summit day, and damaged ropes can combine into a disaster with many deaths.

Recent K2 deaths

Deaths since 2021 have come from falls, avalanches, altitude illness, and rockfall. The record shows:

• 2021: six deaths (including a July avalanche)
• 2022: three deaths (two falls and a suspected altitude illness death)
• 2023: one death
• 2024: three deaths
• 2025: two deaths as of the source’s mid August update (an avalanche and rockfall)

Smaller counts have not made K2 safe, as the 2025 deaths show.

K2 Death Rate vs Everest

This comparison uses one Everest dataset throughout: the Himalayan Database figures reported by Guinness World Records as of February 2025, which list 344 deaths against 13,752 ascents, or about 2.5% (344 ÷ 13,752 = 2.50%). Measured the same way, K2’s ratio of 9.54% is about 3.8 times Everest’s.

Metric	K2	Everest
Height	8,611 m	8,849 m
Recorded deaths	92 (through Dec. 31, 2025)	344 (as of Feb. 2025)
Recorded summits	964	13,752
Death to summit ratio	~9.5%	~2.5%
Main hazards	Technical terrain, Bottleneck serac, storms, rockfall	Altitude illness, Khumbu Icefall, crowding, cold
Rescue environment	Very limited above Base Camp	Limited, but with far more infrastructure

Four cautions apply:

• Different compilations. K2’s figures come from the 8000ers.com, ExplorersWeb, and Alan Arnette compilation, and its cutoff is about ten months later than the Everest figures, so the comparison is approximate.
• Other Everest sources differ slightly. Wikipedia, for example, cites 12,884 summits and 340 deaths through December 2024 (about 2.6%). That is a separate compilation with different counting rules, and it is not used in the table above. Both versions land near 2.5 to 2.6%, so the conclusion does not change.
• Ascents, not people. Both counts record ascents rather than unique climbers, so repeat summiters are counted each time.
• Different denominators. Everest is often quoted at “1 to 2%” or lower, which usually uses a different denominator, such as everyone who attempts. Only compare ratios built the same way.

The larger point is structural. Everest’s high death toll reflects enormous traffic, while K2’s higher ratio reflects small traffic, steeper terrain, and thinner support. Neither number is your personal risk.

K2 vs Annapurna Death Rate

In current Himalayan Database figures, Annapurna I has a higher death to summit ratio than K2’s 9.54%. As of February 2025, the database (as reported by Guinness World Records) lists 75 deaths against 559 ascents, or 13.42%. For comparison, in the same dataset:

• Dhaulagiri: 92 deaths against 707 ascents (13.01%)
• Kangchenjunga: 54 deaths against 690 ascents (7.83%)
• Everest: about 2.5%

One caveat matters here. The Himalayan Database covers peaks in Nepal, and K2 is in Pakistan, so K2’s 9.54% comes from a separate compilation with a different cutoff. Reading across the two sets gives a rough guide, not a precise comparison.

The two mountains are dangerous for different reasons. K2’s risk comes from steep technical terrain, the Bottleneck, and violent weather. Annapurna’s comes largely from avalanche exposure and unstable snow slopes, and it has historically seen fewer summits. Annapurna also appears to have followed the same pattern as K2: older tallies reportedly put it near one death for every three summits, and that ratio fell to about 13% as more climbers reached the top.

Older snapshots, such as a figure of 73 Annapurna deaths from 2024, reflect earlier cutoffs and are not in conflict with the newer numbers. Ratios shift depending on dataset and cutoff date, so treat any “deadliest mountain” claim as a snapshot and not a permanent title.

Are There Still Bodies on K2?

Yes, some climbers who died on K2 remain on the mountain, but no reliable count exists. Recovery above the high camps is extremely difficult, dangerous, and costly, so many bodies stay where they fell or are buried by snow and ice.

Some remains do come to light or are recovered. Wikipedia’s record notes, for example, that the remains of Liliane and Maurice Barrard, missing since 1986, were found years later, and that remains of climbers lost in February 2021 were located that July. The Gilkey Memorial, a stone cairn near Base Camp, honors the many climbers who died on K2.

How Many People Have Summited K2?

About 964 summits had been recorded through 2025 in the compilation used here, with roughly 800 of them after 2000. Older estimates say “about 800,” such as Wikipedia’s rounded figure from August 2023, so counts vary by source and date.

Summit counts matter because they are the denominator in the death rate. They also count ascents rather than individual people, since some climbers have summited more than once, and they generally include high altitude workers as well as clients.

Other Questions Climbers Ask