Mountaineering Deaths by Decade: Historical Trends 1854–2025
On July 14, 1865, Edward Whymper and six companions stood on the summit of the Matterhorn. Four of them died on the descent when one climber slipped and pulled the rest down the north face. Charles Dickens denounced the “fool-hardihood” and “contempt for and waste of human life” of Alpine climbing. The Times asked: “Is it life? Is it duty? Is it common sense? Is it allowable? Is it not wrong?” One hundred sixty years later, the Matterhorn has killed approximately 600 climbers, and Everest 339. This investigation traces the seven distinct eras of mountaineering fatalities — what changed, what didn’t, and why the death rate per climber on the world’s highest mountain has fallen from over 5% pre-1990 to roughly 0.69% in 2015–2025.
deaths 1922–2025
pre-1990 vs 2015–25
deaths since 1865
2010–2025 era
There is no single story of mountaineering fatalities. There are seven. The Golden Age (1854–1865) opened with optimism and closed with the Matterhorn disaster; the Reconnaissance Era (1920s–1940s) sent British expeditions into the unknown Himalaya and saw George Mallory disappear at 28,000 feet; the Heroic Age (1950s–1970s) achieved the first ascents of Everest, K2, and Annapurna at sobering cost; the Pre-Commercial Era (1980s) saw fatalities accelerate as more climbers attempted the peaks; the Commercial Awakening (1990s) gave us the 1996 Into Thin Air storm and the first commercial guided expeditions; the Mass Commercial Era (2000s–2010s) industrialized Himalayan mountaineering and brought average death tolls higher even as per-climber risk fell; and the Modern Calibration Era (2020s) is where the industry now sits — with the lowest normalized risk in history alongside the highest absolute participation. Each era was shaped by specific technologies, specific tragedies, and specific cultural assumptions about what climbers owed each other. This investigation maps the data and tells the story, era by era.
Sources. Cumulative Everest fatality data primarily from Alan Arnette’s 2026 Everest by the Numbers, which draws on the Himalayan Database (Salisbury, Hawley) through end of 2025. Decade-by-decade Everest counts cross-referenced from Grokipedia’s Mount Everest death statistics, Nation Africa reporting (citing Himalayan Database), and Outdoor Action Plan’s historical compilation. Golden Age and Matterhorn data from Wikipedia’s documented first ascent records, the History Cooperative, Project MUSE’s Mountaineering Accidents historical analysis, and Encyclopedia.com. K2, Annapurna, and 8000m historical fatality data cross-referenced from Investigation 08 sources including Wikipedia’s List of deaths on eight-thousanders. Methodology note. Decade boundaries are reporting conveniences — actual fatality patterns don’t change discretely on January 1 of decade-start years. We use calendar decades because the Himalayan Database does, while noting that the “eras” we describe (Golden Age, Reconnaissance, etc.) span across decade boundaries. Caveat on per-climber rates. Death rates “per climber” or “per ascent” can be calculated multiple ways — successful summits only, all climbers above base camp, all expedition members, etc. We use the most commonly cited Himalayan Database figures and note when different denominators apply.
The 170-year trajectory in eight numbers
Before the decade-by-decade walkthrough, here is the cumulative picture as of December 2025:
The master decade table
Below are the Everest fatality counts by decade — the cleanest single dataset showing how the picture has evolved. Other mountains (Matterhorn, K2, Annapurna) follow related but distinct patterns covered in the era-by-era sections that follow.
| Decade | Everest deaths | Climbers attempting | Defining events |
|---|---|---|---|
| 1921–1929 | ~13 | ~50 | British reconnaissance era. 1922: avalanche kills 7 Sherpas (first recorded Everest deaths). 1924: Mallory and Irvine disappear high on the north ridge. |
| 1930–1939 | ~5 | ~80 | Continued British attempts via the north side. Maurice Wilson dies attempting solo summit in 1934. |
| 1940–1949 | 0 | 0 | WWII halts all expeditions. No climbing on Everest. |
| 1950–1959 | ~4 | ~30 | 1953: Hillary and Norgay achieve first ascent. Death rate ~3% in this decade per Himalayan Trekking analysis. |
| 1960–1969 | 6 | ~30 | Continued national expeditions (American, Indian, Swiss). 1963: first American ascent (Whittaker), first ascent of west ridge (Hornbein, Unsoeld) with one death (Bishop). |
| 1970–1979 | 28 | ~120 | Climbing frequency increases sharply. 1975: Junko Tabei first woman to summit. Hannelore Schmatz dies on descent 1979 (her body would remain visible on the mountain for decades). |
| 1980–1989 | 58 | ~280 | Substantial increase in attempts. 1980: Messner solo without oxygen. 1985: Dick Bass becomes first commercial-style client to summit, foreshadowing the commercial era. |
| 1990–1999 | 60 | ~770 | 1996 disaster: 15 deaths in single year, including the 8-death May storm that became Into Thin Air. Commercial expedition era launches. First-time clients pay $65,000+ for guided ascents. |
| 2000–2009 | ~55 | ~2,100 | Massive commercial expansion. 2006: 11 deaths including David Sharp incident. Multiple operators competing; oxygen logistics improve substantially. |
| 2010–2019 | ~110 | ~5,800 | The deadliest decade ever. 2014: Khumbu Icefall avalanche kills 16 Sherpas. 2015: Nepal earthquake triggers avalanche, 22 deaths at Base Camp. 2019: overcrowding/queues, 18 deaths. |
| 2020–2025 | ~50 | ~3,800 | Lowest normalized rate in modern history (~0.69%). 2020–22: COVID disruptions reduce participation. 2023: 18 deaths. 2024: 8–9 deaths (May 21 Hillary Step cornice collapse killed 4). 2025: only 3 deaths despite massive participation. |
Counts and ranges are approximate, reflecting different methodologies across sources. Decade totals are most reliable for 1970+; earlier decades have some uncertainty. Cumulative figure (339) reflects Alan Arnette’s 2026 compilation through end of 2025 season.
Reading this table generates two opposite reactions depending on which column you read first. The death count column suggests mountaineering has become dramatically more dangerous — from ~13 deaths in the 1920s to ~110 in the 2010s. But the climbers-attempting column reveals the opposite: participation grew approximately 80-fold over the same period (from ~50 climbers in the 1920s to ~5,800 in the 2010s). The per-climber death rate has fallen from over 5% pre-1990 to approximately 0.69% in the 2015–2025 decade — a roughly 7-fold improvement in individual climber safety. The mountain has gotten less dangerous per attempt; the absolute count has gotten higher because so many more people are attempting. Both framings are true. The popular media narrative typically emphasizes the first; the data emphasizes the second.
The Golden Age, 1854–1865
Modern mountaineering as a sport begins with Alfred Wills’ 1854 ascent of the Wetterhorn — though the first ascent of Mont Blanc (1786) and various pre-Golden Age climbs preceded it. The eleven years from 1854 to 1865 saw nearly all the major Alpine peaks summited for the first time, predominantly by British climbers accompanied by Swiss or French guides. The Alpine Club of Great Britain was founded in 1857. Mountaineering became a fashionable sport with official guides, increasingly technical equipment, and dedicated literary chronicles.
The era closed dramatically with the Matterhorn. On July 14, 1865, Edward Whymper led a team of seven to the summit. On the descent, one climber slipped and pulled three others to their deaths when the rope broke between them and Whymper. The four who died — Charles Hudson, Lord Francis Douglas, Douglas Hadow, and guide Michel Croz — fell down the north face. Whymper and the Taugwalder guides survived only because the rope between the two groups had broken. The accident generated a public reckoning unlike anything mountaineering had previously seen.
The 1865 Matterhorn descent — 4 deaths
July 14, 1865Whymper’s seven-person team summited the Matterhorn on the morning of July 14, 1865. On the descent, climber Douglas Hadow — inexperienced and exhausted — slipped on icy rocks below the summit. The rope between him and guide Michel Croz pulled Croz off balance; the chain reaction pulled down Lord Francis Douglas and Charles Hudson before the rope broke between Douglas and the next climber (one of the Taugwalder guides). The four fell more than 4,000 feet down the north face. Three bodies were recovered; Lord Douglas’s body was never found. The public reaction was severe — Charles Dickens denounced the “fool-hardihood” of Alpine climbing; The Times asked “Is it not wrong?”; rumors circulated that Whymper had cut the rope (a subsequent inquiry found no evidence). The disaster ended the Golden Age and inaugurated a more sober era of mountaineering.
What changed: from optimism to caution
The Matterhorn disaster reshaped public discourse about mountaineering risk. The Times’s editorial — “Is it life? Is it duty? Is it common sense? Is it allowable? Is it not wrong?” — captures a Victorian framing that mountaineering needed moral justification. The subsequent decades saw climbing techniques mature, professional guide associations form, and an institutional culture develop around risk management. The Matterhorn would go on to kill approximately 600 more climbers in the 160 years since the first ascent — but the era of casual, lightly-equipped, semi-amateur first ascents on major peaks effectively ended in 1865.
The Reconnaissance Era, 1920s–1940s
Within fifteen years of the Matterhorn ascent, the major Alps peaks had been climbed, and mountaineers began looking elsewhere — the Andes, the Caucasus, the Himalaya, the mountains of Africa, the western mountains of North America. Interest in Himalayan climbing began seriously in the 1880s but accelerated after WWI, when the British Mount Everest Committee was formed and the 1921 reconnaissance expedition mapped the approach. The Reconnaissance Era was characterized by national expeditions to peaks that had never been climbed, with limited technology and uncertain logistics — and the death tolls reflected the conditions.
1922 North Col avalanche — 7 deaths
June 7, 1922The first recorded fatalities on Everest. The 1922 British expedition was attempting the north side when an avalanche on the North Col swept seven Sherpa porters to their deaths. These were the first known deaths on the world’s highest mountain and established the pattern — continuing through the 2014 Khumbu Icefall avalanche — that Sherpa fatalities would disproportionately reflect the actual mountain dangers while Western climbers received most of the historical attention.
Mallory and Irvine disappear — 2 deaths
June 8, 1924George Mallory and Andrew Irvine were last seen approximately 240 meters from the summit of Everest on June 8, 1924, climbing strongly into the clouds. They did not return. Whether they reached the summit before dying remains the most enduring mystery in mountaineering history. Mallory’s body was recovered in 1999 with signs of fall trauma; Irvine’s was found in 2024 with new evidence about their final hours. The popular framing — “George Mallory died because he wanted to climb the mountain that was there” — has shaped mountaineering culture for a century.
What changed: the framework for high-altitude mountaineering
The Reconnaissance Era invented high-altitude mountaineering as a coherent practice. Supplemental oxygen, established camps, fixed ropes on technical sections, base camp medical capabilities, and the staged-ascent acclimatization protocol all emerged in this period. The 1953 first ascent of Everest (Hillary, Norgay) was the culmination of three decades of accumulated technique. WWII halted all Himalayan climbing in the 1940s; the resumption in 1950 marked the start of the Heroic Age.
The Heroic Age, 1950s–1970s
The Heroic Age opened in 1950 with the French expedition’s first ascent of Annapurna I — the first 8,000-meter peak ever summited — and effectively closed in 1964 when the last 8000er, Shishapangma, was finally climbed. In 14 years, 12 nations completed first ascents of the 14 eight-thousanders, often at substantial cost. By any reasonable measure, this was the most consequential single period in mountaineering history.
The death tolls reflected the conditions. Annapurna (1950, French) had no fatalities on the first ascent but Maurice Herzog and Louis Lachenal suffered devastating frostbite injuries. K2 (1954, Italian) was achieved without fatalities. Everest (1953, British) was achieved without fatalities, but the 1922 deaths had already established the mountain’s pattern. Other peaks were less fortunate. Nanga Parbat killed 31 climbers before its first ascent in 1953 (Hermann Buhl, solo). Many of the early Himalayan deaths were Sherpas and porters whose names have been historically less well documented than Western climbers’.
Nanga Parbat — 31 deaths before first ascent
1895–1953Before Hermann Buhl made the first ascent of Nanga Parbat in 1953 — solo, in his last hours of consciousness after an unplanned bivouac at 8,000m — the mountain had killed 31 climbers attempting the same goal. The 1934 German expedition lost 10 climbers (4 Germans and 6 Sherpas) in a single storm. The mountain’s nickname “Killer Mountain” predates the modern fatality data by decades. See Investigation 08 for the full Nanga Parbat fatality profile, which remains at approximately 20% to this day.
What changed: 8000m climbing became a known endeavor
Before 1950, no one had reliably summited a peak above 8,000 meters. By 1964, all 14 had been climbed. The Heroic Age transformed 8000m mountaineering from theoretical possibility to established practice — with documented routes, known hazards, base camp infrastructure, and the cultural framework that would carry forward into the commercial era. The death-zone physiology that Investigation 07 covers in detail was first understood in this period; the techniques that Reinhold Messner and others would later use to climb without oxygen were pioneered in the Heroic Age.
The Pre-Commercial Era, 1980s
The 1980s were a transitional decade. Reinhold Messner summited Everest without supplemental oxygen in 1978 with Peter Habeler — and the following year became the first person to climb all 14 eight-thousanders (1986). The decade saw 58 Everest deaths against approximately 280 climbers attempting — a much higher absolute count than the 1970s but with a per-climber rate that remained roughly stable.
The Pre-Commercial Era ended with a specific event in 1985: Dick Bass, a Texas oil and ski-resort executive with limited mountaineering experience, summited Everest as the first commercial-style client. Bass paid for guides, used established infrastructure, and demonstrated that Everest could be climbed by individuals whose primary qualification was paying for the expedition rather than years of elite mountaineering experience. This single ascent foreshadowed the entire commercial era that would dominate the 1990s and beyond.
What changed: the demographic shifted
Through the 1970s, Everest and other 8000m climbers were overwhelmingly drawn from a small pool of elite mountaineers, often selected through national-expedition processes that screened for technical climbing experience and high-altitude track record. The 1980s opened the door to commercial clients — climbers who could afford expedition costs but had less elite background. Dick Bass’s 1985 ascent was the proof of concept; the next decade would scale it.
The Commercial Awakening, 1990s
The 1990s reshaped Himalayan mountaineering. Commercial guided expeditions on Everest emerged as a recognized industry, led initially by Rob Hall’s Adventure Consultants and Scott Fischer’s Mountain Madness, joined by others through the decade. Climbers paid approximately $65,000 (1996 dollars; equivalent to ~$130,000 today) for guided ascents with substantial Sherpa support, oxygen logistics, and established camps.
The decade also produced the most famous disaster in Everest history: the May 10–11, 1996 storm that killed 8 climbers in a single day, including Hall and Fischer themselves. Jon Krakauer’s Into Thin Air made the event globally known and shaped public understanding of commercial Everest mountaineering for an entire generation. The total 1996 death count on Everest was 15 — the deadliest single year to that point.
The 1996 Everest disaster — 8 deaths (15 total for the year)
May 10–11, 1996Multiple commercial expeditions summited on May 10, 1996, in conditions that deteriorated rapidly through the afternoon. A severe storm hit the upper mountain that evening, trapping climbers in the death zone overnight. Rob Hall (Adventure Consultants founder), Scott Fischer (Mountain Madness founder), Andy Harris (guide), Doug Hansen, and Yasuko Namba died in the storm and its aftermath; three Indian climbers from a separate expedition died high on the north side the same day. Three more climbers died elsewhere on Everest in 1996, bringing the year’s total to 15 — the deadliest year on record until 2014. The disaster prompted multiple books and films and fundamentally shaped public understanding of commercial Everest mountaineering. Jon Krakauer’s Into Thin Air (1997) remains the most-read mountaineering book in publishing history.
What changed: commercial mountaineering had to grow up
The 1996 disaster forced the commercial Everest industry to mature rapidly. Operator-published turnaround time policies, more rigorous oxygen logistics, dedicated weather forecasting, expanded Sherpa-to-client ratios, and stricter client vetting all emerged in the years following 1996. The per-climber death rate began declining shortly afterward — partly through these operator improvements, partly through equipment advances, and partly because the commercial industry consolidated around fewer, more experienced operators. The era’s lesson — captured in Investigation 03 — that operator quality dominates outcome variance, was learned the hardest possible way.
The Mass Commercial Era, 2000s–2010s
The 2000s and 2010s scaled commercial Himalayan mountaineering by an order of magnitude. Everest climber counts grew from ~2,100 in 2000–2009 to ~5,800 in 2010–2019. Nepali-operated commercial expeditions emerged alongside the established Western operators, often at substantially lower price points. Cho Oyu and Manaslu became standard “first 8000m” peaks. The 14 eight-thousanders climbed by all-summit climbers became a recognized achievement — though one that, as Investigation 08 details, remains rare and lethal.
The 2010s were the deadliest decade on Everest in absolute terms, with approximately 110 fatalities. Three specific events dominated the toll.
2014 Khumbu Icefall avalanche — 16 deaths
April 18, 2014An ice serac collapsed in the Khumbu Icefall, sweeping a group of Sherpa porters who were carrying equipment to the upper camps. 16 Sherpas died in a single event — at the time, the deadliest single incident in Everest history. The disaster prompted protests by Sherpa communities and a partial shutdown of the climbing season. It also contributed to the Khumbu Glacier instability documented in Investigation 12.
2015 Nepal earthquake — 22 deaths at Base Camp
April 25, 2015The magnitude 7.8 Nepal earthquake triggered an avalanche from Pumori that swept through Everest Base Camp. 22 climbers and Sherpas died at Base Camp itself, including Google executive Dan Fredinburg. The broader Nepal earthquake killed approximately 9,000 people across the country. The 2015 Everest climbing season was cancelled entirely — the first such cancellation since the Sherpa protests of 2014.
2019 overcrowding deaths — 11 above-summit fatalities
May 2019The 2019 climbing season produced 11 fatalities directly attributable to summit-day queueing and time-budget exhaustion. The iconic Nirmal Purja photograph of climbers stacked along the summit ridge crystallized the public understanding of how commercial Everest had reached operational limits. The 2019 season generated the most substantive policy discussion about Everest climbing limits in two decades — though the actual reforms (the September 2025 permit increase from $11,000 to $15,000 and proposed prior-7,000m-experience requirements) came years later.
What changed: the per-climber rate kept falling despite higher absolute counts
The 2010s’ 110-death toll on Everest looks catastrophic in absolute terms. But during the same decade, climber participation grew so rapidly that the per-climber death rate continued declining. The 2014 Icefall avalanche (16 Sherpa deaths in one event) and 2015 earthquake (22 deaths) account for nearly 35% of the entire decade’s deaths in two single events — both largely unrelated to climbing decisions in the conventional sense. Outside those events, the underlying per-climber risk continued to fall. This pattern is important to understand: the absolute count is dominated by major events, while the per-climber rate reflects the underlying operational quality of commercial expeditions.
The Modern Calibration Era, 2020s
The 2020s opened with the COVID-19 pandemic, which substantially reduced Everest participation in 2020 and 2021. By 2022, participation had recovered to pre-pandemic levels. The 2023 season produced 18 fatalities — a high count but reflecting overcrowding and equipment failures more than systemic operator failure. 2024 produced 8–9 deaths, including the May 21 Hillary Step cornice collapse that killed 4 climbers in a single event. 2025 produced only 3 deaths against well over 1,000 climbers on the mountain — one of the safest seasons in modern history.
The cumulative picture: the 2015–2025 decade Everest death rate is approximately 0.69% — the lowest sustained rate in Himalayan mountaineering history. By comparison, K2’s historic rate hovered between 12% and 25% depending on era, and Annapurna’s modern rate is approximately 13–20%. Everest in 2026 is, per the data, statistically safer than it has ever been despite being more crowded than ever.
2024 Hillary Step cornice collapse — 4 deaths
May 21, 2024On May 21, 2024, a cornice collapsed at the Hillary Step (the famous final-ridge section just below Everest’s summit), sweeping four climbers to their deaths. The event accounted for nearly half of all 2024 Everest fatalities in a single incident. The Hillary Step has been progressively changing over the past decade — partly from glacier and snow-cornice changes connected to broader Himalayan warming patterns documented in Investigation 12. The 2024 event raised serious questions about whether the Hillary Step remains structurally stable in its modern form.
What changed: the per-climber rate hit its lowest point
The 2020s have produced the lowest sustained per-climber death rate in Everest history. The combination of mature commercial operator infrastructure, dedicated weather forecasting, improved oxygen logistics, Sherpa expertise accumulated over 30+ years of commercial-era experience, premium-tier operator quality differentiation (per Investigation 10), and effective rescue/helicopter capability have all compounded into the safest operating environment in the mountain’s history. The absolute counts remain high because so many climbers attempt — but each individual climber today faces meaningfully lower risk than each climber faced 20, 40, or 60 years ago.
What the 170-year picture tells us
Pulling back from the era-by-era walkthrough, five patterns emerge across the full historical sweep of mountaineering fatalities.
1. Per-climber risk has fallen substantially across all eras
From the Matterhorn descent (4 deaths out of 7, ~57% fatality rate on that single climb) through the Reconnaissance Era (~10% per attempt on early Everest expeditions) through pre-1990 Everest (~5%+ death rate) through the 2015–2025 decade (~0.69%), the trend is unambiguously toward lower per-climber risk. The mountains have not gotten less dangerous; the operational quality of expeditions has gotten substantially better.
2. Absolute death counts have risen because participation has risen faster than safety has improved
The 1920s saw ~50 Everest climbers and ~13 deaths. The 2010s saw ~5,800 climbers and ~110 deaths. The participation grew 116-fold; the deaths grew 8-fold. Per-climber risk improved by approximately a factor of 14, but the absolute count rose because the denominator grew so much. Public framings that focus on “Everest deaths are at record highs” without contextualizing participation miss the structural improvement in individual climber outcomes.
3. Major disasters dominate decade-level statistics
The 2014 Khumbu Icefall avalanche (16 Sherpa deaths) and 2015 earthquake (22 Base Camp deaths) together account for 35% of the 2010s’ decade total. The 1996 disaster (15 deaths) accounted for 25% of the 1990s’ decade total. Single events disproportionately shape the decade-level numbers. The underlying per-climber risk in years without major events is meaningfully lower than the decade averages suggest.
4. Sherpa fatalities have been disproportionately undercounted historically
Of the 339 cumulative Everest deaths through 2025, 132 are hired workers (primarily Sherpas) — approximately 39% of all fatalities. In the early decades (1920s-1950s), Sherpa fatalities were often documented only by count and rarely by name. The 2014 Khumbu Icefall avalanche, which killed 16 Sherpas in a single event, was an inflection point in public attention to Sherpa risk; the response included partial-season closures, increased Sherpa life insurance requirements, and substantially improved compensation. The cumulative pattern is that Sherpas have borne a disproportionate share of the mountain’s risk relative to the cultural attention given to Western climbers — a pattern that persists in attenuated form to this day.
5. The descent-vs-ascent fatality pattern has shifted
The 1921–2006 historical pattern (62% of deaths on ascent, 38% on descent — Huey et al. 2020) has shifted to a 2022–2023 modern pattern (56% of above-base-camp deaths on descent — Himalayan Database via Zara Tours analysis). This shift likely reflects improvements in ascent-phase logistics and oxygen reliability that haven’t equally improved descent-phase outcomes — climbers are reaching the summit more often, but the descent remains the relatively under-managed phase of the climb. The 2014 PLoS One finding that 61.7% of all deaths occurred after summiting (vs 46.9% in the prior sample) reinforces this shift. See Investigation 07 for the full death-day failure analysis.
Frequently Asked Questions
How many people have died on Mount Everest?
As of December 2025, 339 people have died on Mount Everest since the first British expedition in 1922 (Alan Arnette 2026 / Himalayan Database). Of these, 207 are members (foreign climbers, often paying clients) and 132 are hired workers (primarily Sherpas). The mountain has averaged about 5 deaths per year historically, rising to roughly 7 deaths per year from 2010 to 2025 as participation has grown. Approximately 200–250 bodies remain on the mountain, mostly in the death zone above 8,000m where recovery is dangerous, expensive, and often impossible.
What was the deadliest year on Everest?
By single-year absolute count, 2015 was the deadliest with 22 deaths at Base Camp from the avalanche triggered by the Nepal earthquake. 2014 was the second-deadliest with 19 deaths, including 16 Sherpas killed in the Khumbu Icefall avalanche. 1996 was the third-deadliest with 15 deaths, including the famous May 10–11 storm that killed 8 in a single event (the Into Thin Air disaster). 2019 produced 18 deaths primarily attributed to overcrowding and summit-day queueing. By per-climber rate, the earliest decades were much deadlier — the pre-1990 era had over 5% death rate among climbers, compared to approximately 0.69% in the 2015–2025 decade.
Has Everest become safer or more dangerous over time?
Both, depending on which framing you use. Per-climber risk has fallen substantially — from over 5% pre-1990 to approximately 0.69% in 2015–2025, a roughly 7-fold improvement. Absolute death counts have risen — from ~13 deaths in the 1920s to ~110 in the 2010s — because participation grew approximately 100-fold over the same period. The mountain has not gotten less dangerous; the operational quality of commercial expeditions has gotten substantially better. Public narratives typically emphasize the absolute count (“record deaths on Everest”) because the count is more dramatic; the per-climber rate tells a different story about the structural improvement in commercial mountaineering safety since 1996. Both framings are accurate; both matter; neither alone is the full picture.
What was the 1865 Matterhorn disaster?
On July 14, 1865, Edward Whymper led the first ascent of the Matterhorn — a 7-person team including three British climbers, one Italian guide (Michel Croz), and two Swiss guides (Peter Taugwalder and his son). On the descent, Douglas Hadow (the most inexperienced climber) slipped on icy rocks, pulling guide Michel Croz, Lord Francis Douglas, and Charles Hudson down the north face. The rope between them and the next climber broke; the four fell more than 4,000 feet. Three bodies were recovered; Lord Douglas’s body was never found. Whymper and the Taugwalder guides survived. The event is considered to mark the end of the Golden Age of Mountaineering (1854–1865) and triggered a substantial public reckoning about Alpine climbing risk, including denunciations from Charles Dickens and The Times of London. The Matterhorn has since killed approximately 600 climbers and continues to claim 8–10 lives per year.
When did commercial Everest expeditions start?
The commercial era is conventionally dated to Dick Bass’s 1985 summit — the first time a paying client without elite mountaineering background reached the top of Everest. The 1996 disaster (8 deaths in a single storm, including the founders of two leading commercial operators) marked the maturation of commercial mountaineering as a recognized industry that needed to operate professionally. Adventure Consultants (founded 1991 by Rob Hall), Mountain Madness (founded 1985 by Scott Fischer), and International Mountain Guides (founded 1986) were among the early Western commercial operators on Everest. The Nepali-operated commercial expedition industry — Seven Summit Treks, Imagine Nepal, 8K Expeditions, others — emerged primarily in the 2000s and 2010s, often at substantially lower price points than the Western operators (see Investigation 10 for the modern operator landscape).
Why does the Matterhorn kill so many climbers?
The Matterhorn — approximately 600 cumulative deaths since 1865, 8–10 per year currently — has several specific factors that compound its danger. (1) Sheer climber volume: up to 3,000 climbers attempt the mountain each season, with up to 130 attempting the summit via the Hörnli Hut on peak days. (2) Variable difficulty perception: the climb is non-technical relative to other Alps 4,000m peaks but is meaningfully harder than guidebooks often suggest, and many climbers underestimate it. (3) Weather exposure: the long, exposed ridge has minimal shelter and rapid weather changes can trap climbers high on the mountain. (4) Permafrost and rockfall: as documented in Investigation 12, the Matterhorn has had multiple major rockfall events as permafrost thaws — including the 2003 event that evacuated 90 climbers. The Tomb of the Unknown Mountaineer at the Zermatt mountaineering cemetery commemorates the 500+ deaths the mountain has caused since 1865; several dozen climbers’ bodies have never been found.
How has the descent-vs-ascent fatality pattern changed?
The historical pattern (1921–2006) showed 62% of Everest deaths occurring during ascent or route preparation and 38% during descent (Huey et al. 2020 PLoS One analysis). The modern pattern (Himalayan Database 1922–2023) shows 56% of above-base-camp fatalities occurring during descent — a substantial shift. The 2020 PLoS One follow-up study found that 61.7% of recent (2006–2019) Everest fatalities occurred after summiting, even though descent is a small fraction of total time on the mountain. The shift reflects improvements in ascent-phase logistics (oxygen, weather forecasting, fixed ropes, Sherpa support) that haven’t equally improved descent-phase outcomes. Climbers are reaching the summit more reliably than ever, but the time-budget exhaustion, oxygen depletion, and physiological deterioration that compound on the descent remain the under-managed phases of the climb. See Investigation 07 for the detailed analysis of summit-day failure modes.
Are Sherpa fatalities counted in the Everest death total?
Yes — the Himalayan Database explicitly tracks both “members” (foreign climbers, often paying clients) and “hired workers” (primarily Sherpas plus other ethnic groups including Tamang, Magar, and Rai). Of the 339 cumulative Everest deaths through 2025, 132 are hired workers — approximately 39% of all fatalities. In the earliest decades, Sherpa fatalities were often recorded only by count rather than by name; modern documentation is more comprehensive. The 2014 Khumbu Icefall avalanche (16 Sherpas killed in a single event) was an inflection point in public attention to Sherpa risk, and prompted substantial improvements in Sherpa life insurance requirements, compensation, and family support systems. The cultural attention given to Western Everest climbers historically substantially exceeded their share of the mountain’s fatalities — a pattern that the modern industry has partially but not fully addressed.
The 170-year history of mountaineering deaths is a story of two parallel trends running in opposite directions. Per-climber risk has fallen dramatically across every measurable interval — from the Matterhorn descent’s catastrophic ratios through the Reconnaissance Era’s ~10% per-attempt rate through pre-1990 Everest’s 5%+ death rate down to the modern 0.69% rate of the 2015–2025 decade. Absolute death counts have risen dramatically over the same period because participation has grown faster than safety has improved. Both trends are real. Both matter. The popular framing — “Everest deaths are at record highs” — emphasizes the absolute count and misses the structural improvement; the industry framing — “Everest is statistically safer than ever” — emphasizes the per-climber rate and risks normalizing the absolute scale of loss. The honest framing combines both: commercial mountaineering has become substantially safer per attempt, has scaled to a point where each year produces approximately 7 deaths on Everest alone, has structurally redistributed risk (with descent now claiming a greater share than ascent), and continues to disproportionately rely on Sherpa labor whose fatalities have historically been undercounted in public attention. The mountains haven’t changed in 170 years. The climbers have. The infrastructure has. The expectations have. And — slowly, era by era — the operational practices that determine who comes home have gotten meaningfully better.
Sources and Verification
This investigation was built primarily from the foundational Everest fatality dataset, peer-reviewed analysis, and historical mountaineering literature:
- The Himalayan Database (Salisbury, Hawley) — the foundational dataset for all Nepal/Tibet expedition records dating back to 1905. Source for all cumulative summit and fatality counts.
- Alan Arnette’s Everest by the Numbers: 2026 Edition (January 2026) — the canonical annual synthesis of Himalayan Database data. Source for the 339 cumulative deaths (207 members + 132 hired workers), the ~5/year historical and ~7/year recent averages, and the 0.69% normalized death rate for 2015–2025.
- Grokipedia / Wikipedia: List of Mount Everest death statistics — for decade-by-decade death counts and the 5%+ pre-1990 vs 1% modern death rate comparison.
- Huey, Salisbury, Wang, Mao (2020) — PLoS One follow-up analysis of 5,800+ Everest climbers, including the descent-vs-ascent fatality phase analysis and the finding that 61.7% of recent fatalities occurred after summiting.
- Nation Africa reporting — for the decade-by-decade breakdown (1970s: ~28 deaths; 1980s: ~58; 1990s: ~60; etc.).
- Outdoor Action Plan and Himalayan Trekking — for historical era framing and decade trajectory.
- Glorious Himalaya — for the early-era (1920s-50s) characterization and the specific incident summaries (1922 avalanche, 1996 storm, 2015 earthquake, 2019 overcrowding).
- Zara Tours / Mount Everest Deaths analysis — for the modern 56% descent-share statistic and 200–250 bodies remaining on the mountain.
- Wikipedia: First Ascent of the Matterhorn — for the 1865 disaster details (Whymper, Croz, Hudson, Douglas, Hadow, Taugwalders).
- Encyclopedia.com, History Cooperative, Project MUSE Mountaineering Accidents — for Golden Age period framing (1854 Wetterhorn → 1865 Matterhorn) and Victorian public-reaction documentation.
- The Canadian Encyclopedia — for the post-Matterhorn expansion of mountaineering to global ranges.
- Switzerland Highlights: How Dangerous is the Matterhorn Today — for the 600 cumulative Matterhorn deaths since 1865 and the current 8–10 deaths/year rate.
- Investigation 01 of this series (Everest Death Map) — for the detailed 1922–2025 Everest fatality map and the interactive geographic display.
- Investigation 07 of this series (Everest summit day deaths) — for the death-zone risk patterns and the death-day failure mode framework.
- Investigation 08 of this series (8000ers ranked) — for K2, Nanga Parbat, Annapurna fatality data cross-references.
- Investigation 12 of this series (Glacier recession) — for the Khumbu Icefall instability and Hillary Step cornice changes context.
Methodology and caveats. Decade boundaries are reporting conveniences; the “eras” we describe (Golden Age, Reconnaissance, etc.) span across decade boundaries. Death counts vary slightly across sources due to inclusion criteria (base camp vs above; members only vs all hired workers; etc.); we have used the most commonly cited Himalayan Database figures throughout. Per-climber rates can be calculated multiple ways — we use Alan Arnette’s 2026 normalized rate methodology for the 0.69% figure and note when different denominators apply. Right of response. Researchers or operators with documented updates to decade counts or era characterizations are invited to contact our editorial team for incorporation in the November 2026 update.
Published May 21, 2026 · Data through 2025 season · Next scheduled review: November 2026
Related Investigations
Subscribe to the Mountain Planning Brief
Get notified when each new investigation publishes and when this article gets its annual data refresh after the next climbing season concludes. One email per release. No marketing. Just the work.