10 facts about Everest success and death rates, based on scientific data

Today’s post is all about an exciting new scientific research paper. No, no no… don’t click the back button just yet – I promise you this one’s really interesting.

Once a year (except this year, obviously), there is an Everest feeding frenzy as traditional and social media sink their teeth into the latest Everest season, producing an avalanche of opinion about how overcrowded and easy Everest is to climb these days.

Every year, photos of queues and soundbites from climbers, both on the mountain and watching from home, are cited as evidence the mountain is too overcrowded, climbers are too inexperienced, and that climbing Everest is simultaneously becoming too easy and too dangerous.

Climbers on Everest's north ridge
Climbers on Everest’s north ridge

Barring a few lone voices, such as the excellent Alan Arnette whose annual Everest coverage has become the unrivalled source of contemporary Everest history and commentary, rarely does anyone delve into the data to try to connect opinion with reality.

Which is why I was very excited to see a paper entitled Mountaineers on Mount Everest: Effects of age, sex, experience, and crowding on rates of success and death published on the open-access scientific journal PLOS ONE last week.

Written by Raymond B. Huey, a biologist from the University of Washington, Cody Carroll and Jane-Ling Wang, both statisticians from the University of California, and Richard Salisbury, the technical brains behind Elizabeth Hawley’s Himalayan Database, the paper’s title speaks for itself. How much does overcrowding, the climber’s experience, age and sex affect how likely they are to reach the summit or die trying?

There’s no longer any need to speculate – here are the numbers.

Using the Himalayan Database as their source, the scientists carried out a similar study a few years ago of 2,211 climbers making their first attempt on Everest between 1990 and 2005. Since 2005, the number of people attempting Everest has continued to increase. Using the expanded dataset of 5,324 climbers making their first attempt between 1990 and 2019, the scientists wanted to see if their conclusions were still the same or if trends were changing.

To make the comparison more reliable, they removed Sherpas and other high-altitude workers from their analysis, as well as climbers on non-standard routes, those climbing outside the normal spring season, and data from the years 2014 and 2015 when two significant events (a strike by high-altitude workers and an earthquake) had a dramatic effect on summit success rates.

Here are 10 significant findings from their study that may or may not surprise you.


1 Summit success is becoming more likely

The summit success rates have increased significantly in the last 20 years. In fact, the success rates for 2006–2019 are essentially double those for 1990–2005.

Between 1990 and 2005, 32.7% of 214 women and 32.9% of 1,702 men attempting Everest reached the summit.

Between 2006 and 2019, 68.2% of 548 women and 64.4% of 2,860 men attempting Everest reached the summit.

In other words about two thirds of climbers who went above base camp between 2006 and 2019 reached the summit.

Success and death rates on Everest over time (Graph: Huey/Carroll/Salisbury/Wang)
Success and death rates on Everest over time (Graph: Huey/Carroll/Salisbury/Wang)

This is not really surprising. Commercial operators have been getting more experienced and better equipped every year. Climbers are using more oxygen. There is better weather forecasting, and there are teams led by guides, leaders and Sherpas with multiple Everest expeditions under their belts.


2 Women are more likely to summit and less likely to die, but not by much

As you saw in the last section, between 2006 and 2019, 68.2% of women reached the summit vs. 64.4% of men.

Similarly, between 2006 and 2019, only 0.5% of 548 women vs. 1.1% of 2,860 men attempting Everest did not return home alive.

This may lead us to conclude that women are more likely to reach the summit and less likely to die, but hold your horses. According to the scientists ‘women and men had nearly identical rates of success’. This is because they were using a Wilcoxon signed-rank test, p = 0.9168, two sided.

If you’re wondering WTF that means, so am I, but I think it means the difference between the two stats is not significant (Edita drank her glass of red wine more quickly than I did last night, but that doesn’t mean she’s quicker at drinking red wine than I am – in fact, the reverse is true).


3 Success rates plummet after age 40

If you want to see a graph that drops away as steeply as the South-East Ridge on its plummet to the South Col, feast your eyes on this one:

Everest summit success by age (Graph: Huey/Carroll/Salisbury/Wang)
Everest summit success by age (Graph: Huey/Carroll/Salisbury/Wang)

In both the period 1990–2005 and 2006–2019, summit success rate is more or less flat until age 40, then drops off dramatically. For the period 2006–2019, the rate declines by 1.1% for every year in age. I climbed Everest at just the right time (I was exactly 40 and so was Edita).

More specifically, in the period 2006–2019, climbers older than 59 summited only half as often as younger (<59) climbers (33.3% vs. 63.7%).

On a more positive note, the age data also backs up point 1, that summit success is becoming more likely for all age groups. You can see this by the enormous gap between the red and blue lines in the above graph. For climbers younger than 40, the rate has almost doubled (from 36.0% to 69.1%), and for those over 40 it has leapt from 27.6% to 57.4%.


4 Previous experience at high altitude counts

The Himalayan Database records all of a climber’s experience on peaks over 7,000m in Nepal, and certain significant peaks over 6,000m such as Ama Dablam. It doesn’t record their experience in other high mountain ranges like the Karakoram or Andes, and nor does it record their technical experience on lower mountains.

Nevertheless, this gave the researchers some scope for assessing the experience levels of climbers attempting Everest for the first time. And the results were significant.

In the period 1990–2005, climbers with previous experience on at least one other high Nepalese peak had higher success rates than those without (40.9% vs. 28.1%).

In the period 2006–2019, climbers with previous experience on at least one other high Nepalese peak also had a slightly higher success rate than those without (67.6% vs. 61.5%).

As the researchers point out, this could illustrate the obvious benefits of experience, but it could also reflect an element of self-selection. Not everyone performs well at high altitude, but climbers who have been there before will know whether they’re any good. Climbers who haven’t been to high altitude before, won’t know how well they perform at those heights, so there is a higher chance of some of them being poor at high altitude.


5 Experience matters less now than it used to

The researchers speculate whether improving success rates are due to greater climber experience, but anyone who has been following Everest in recent years will suspect that this is not the case.

In fact, a corollary of points 1 and 4 above is that previous experience at high altitude is not as important to summit success as it used to be.

Those attempting Everest for the first time in the period 2006–2019 had proportionately less experience than those who attempted it for the first time in the period 1990–2005 (30.7% vs. 37.5%). Yet as we know from point 1 above, a greater proportion of them succeeded.


6 Climbers are more likely to succeed from the south side than the north side

Those attempting Everest for the first time in the period 2006–2019 were significantly more likely to succeed from the Nepal side than the Tibet side (65.8% vs. 58.4%).

There could be many reasons for this. The south side is a snow plod, while the North-East Ridge is more of a tricky rock scramble. Camps are higher on the north side, with a main summit assault camp at 8,200m vs. 7,950m on the south side. The north side is drier, colder and windier. More likely, the north side is just harder.

I’d like to think the latter reason is most likely. My summit day from the north side took 18 hours, and it’s comfortably the toughest thing I’ve ever done. But you’d need to ask someone who’s summited from both sides for a more accurate assessment.


7 Everest is becoming safer

Overall death rates have declined between 1990–2005 and 2006–2019, though only slightly (1.6% vs. 1.0%).

Death rates are slightly lower for women than for men (0.5% of 548 women vs. 1.1% of 2,860 men for the period 2006–2019).

Death rates are slightly higher on the north side than on the south side (1.5% vs. 0.7%).

Interestingly, death rates are also independent of previous experience at high altitude (there is no figure to support this in the article, though we could infer it from points 5 and 7. The authors say this observation is consistent with other research papers they have cited).

If all this sounds promising, I should point out that ‘safer’ is not the same thing as ‘safe’, and 1% is still quite a high death rate for a climbing holiday.


8 The older you are, the more likely you are to die

As in life, also on Everest. If the COVID-19 pandemic has made you fed up of staring at exponential graphs, then apologies – here’s another one.

Everest death rates by age (Graph: Huey/Carroll/Salisbury/Wang)
Everest death rates by age (Graph: Huey/Carroll/Salisbury/Wang)

As you can see, both the red graph for 1990–2005 and the blue one for 2006–2019 curve like a banana and threaten to shoot off the top the older you get.

For the period 2006–2019, climbers over 59 have a significantly higher chance of dying than those who are younger (4.1% vs. 0.9%).

On the plus side, as you can see by the dotted line on the graph, death rates have been decreasing for all ages, and on average the death rate has been postponed by 11½ years between 1990–2005 and 2006–2019 – or to put it another way, 60-year-old climbers in the second sample had the same chance of dying as 49-year-old climbers in the first one.

But once again, it’s my duty to put a dampener on things. If you’re sitting at home thinking of climbing Everest in your retirement, then 4.1% is still a bit sobering.


9 Overcrowding has increased significantly in recent years

This will come as no surprise to anyone who reads the tabloid press. Last year there was something of a frenzy when Nirmal Purja posted THAT photo of queues on the Hillary Slope.

Everest's most crowded summit days by year (Graph: Huey/Carroll/Salisbury/Wang)
Everest’s most crowded summit days by year (Graph: Huey/Carroll/Salisbury/Wang)

Here’s another graph that resembles the number of porkies uttered by leaders of the free world over the course of history.

The dots on the graph show the number of people leaving for the summit on the busiest day of the year, and the blue line shows the general trend. The dot in the top right represents 23 May 2019, when a staggering 396 people left for the summit, which makes Into Thin Air seem like a quiet night in the Gobi Desert.

There are several reasons for this. The most obvious one is that more people are attempting Everest, but the most significant ones are better weather forecasts and better communication. Weather forecasts are so accurate now that everyone tries to crowd into the best weather day to increase their chance of success.

But it’s not just the busiest day that’s getting busier. The researchers state that the average number of climbers making summit bids above high camp has increased four-fold between 1990–2005 and 2006–2019.


10 Overcrowding does not have a significant effect on either success or death rate

This is perhaps the most surprising finding from this study, and a strong argument against much of the media rage about Everest expeditions, which has focused on overcrowding more recently.

Are photographs of climbers queuing on the Hillary Step evidence that Everest is becoming more dangerous?

It’s certainly true that spending an additional hour or two in queuing traffic over 8,000m increases the risk of altitude sickness and frostbite. It’s therefore reasonable to assume that crowding like this increases death rates and lowers success rates.

But here’s the really interesting thing: the data suggests that it doesn’t.

In 2019, 65.9% of all summit bids were made on just two days, 22 and 23 May. Neither the probability of summiting (92.4% on these two days vs. 90.5% on less crowded days) nor the probability of dying (1% on these two days vs. 1.2% on less crowded days) was significantly different.

It was a similar story in 2018. In that year, 58.2% of all summit bids were made on four days, 16–19 May. In this case the probability of summiting was slightly less on these four more-crowded days (91.1% vs. 95.3% on less crowded days) but not significantly so. In 2018 there were no deaths during summit bids on either crowded or uncrowded days.

This may seem counter-intuitive, but what this tells us is that although Everest is becoming more crowded, it is not yet so crowded that climbing Everest is becoming more dangerous. If you think about it another way, there’s a reason everybody climbs on these more crowded days – because the weather is much better and the mountain is consequently safer. On Everest, weather still poses a much greater risk than people, and as long as operators are able to manage the risk, such as by taking more oxygen and preparing climbers for a longer day, then it’s still better to climb during more crowded days than during marginal weather windows.



In their executive summary, the authors state that their findings can inform prospective climbers on their odds of success or death, and inform governments of the safety consequences and economic impacts of better regulation.

While I’m sure that officials in Nepal’s Ministry of Tourism are busy fashioning policy on useful papers like these, recent history suggests that successive government ministers prefer media soundbites to scientific evidence (a trend that appears to be spilling into more developed countries too).

On a more positive note, we can be more hopeful that prospective climbers are reading this. Since 2014, I also believe that Everest reporting by mainstream media has improved. I’m therefore hopeful that media outlets might pick up on it. If either of these are you then there are clear takeaways.

Climbing Everest is becoming more within the reach of ordinary people like you and me. But it is also becoming more crowded. While it’s not yet reached the stage where crowds have become unmanageable, there will come a time when things reach a tipping point.

My book: if you still haven't bought it yet, now's your chance

Seven Steps from Snowdon to Everest is all about an ordinary guy treading a path that other ordinary people can follow to the top of the world. A key theme is that climbing Everest is as much a journey as a destination. It’s a book that every aspiring Everester and armchair enthusiast should read, and it’s also good fun.

Data suggests that experience at high altitude improves your chances of success. But just as importantly it helps you to enjoy the experience more. Embark on the journey, but enjoy the experience, and don’t leave it too late.

To receive email notifications of my blog posts about mountains and occasional info about new releases, join my mailing list and get a free ebook.
Note: I get a very small referral fee if you buy a book after clicking on an Amazon link.

14 thoughts on “10 facts about Everest success and death rates, based on scientific data

  • September 2, 2020 at 7:02 pm

    That was fascinating, Mark!

  • September 2, 2020 at 9:07 pm

    Thanks for this excellent review of the original paper, Mark, ( which, I suspect, most of us won’t bother to read). At 65, my climbing days are obviously long gone, but my excuse for not having a crack can now go from, “it’s too commercial and crowded” to, “I am likely to die”!
    Having found your blog site in the last year or so I really enjoy your work on it and thank you for providing such interesting reviews and videos etc for the the rest of us.

  • September 2, 2020 at 9:20 pm

    Heehee, thank you for the kind words, Matthew. I’m glad you like the blog, and I’m very happy to provide the dual purpose of both encouraging people to climb Everest and giving them an excuse not to 😉

  • September 3, 2020 at 6:41 am

    Interesting, thank you, again.

    The quality of the operators certainly play a big role. How else one Austrian operator from the harder North side could have put all of their clients on the summit 4 years in a row, while some groups have success rates hovering at 15% (nationality withheld by request)?

    There is a simple explanation for women being slightly more successful than men, even if it is not statistically significant: men are more likely to just go and try less prepared and overconfident. Women are more likely to take it more seriously and not go unless well prepared and ready to prove themselves. After all they do start the endeavour from a slight physical disadvantage.

  • September 3, 2020 at 5:26 pm

    Another great post Mark -you are up there with Alan. -I love Alan, just watching random expedition videos and seeing how he chooses to interact with people makes me a fan before his posts -you know the say the true character of a person is how someone treats a person service workers and definitely how they treat someone they don’t need anything from -and Alan certainly passes that test.

    If you remove the Everest data on avalanche and serac fall deaths in the statistical analysis (which they would have mostly because almost all of these deaths happened to Sherpa guides and professional guides and alternate routes by professionals) you will discover that amateur and guided climbers have died in basically 2 known ways rather evenly. That being health issues (HAPE, HACE, Heart Attack, Stroke etc) and falls. -and when you consider that it is likely that health events caused many if not most of the falls that would note be recorded as a heart attack or stroke or HAPE/HACE because the body was not recovered after the fall or the secondary medical designation would not be necessary as the actual cause of death was certainly the fall, it would completely explain why the death statistics remain steady no matter the crowd size is on the mountain.

    The data basically concludes that 1% of the population has underlying health vulnerabilities not suited for high altitude. And knowing this means, in the future, a medical pre screening that includes genetic factors and pulmonary/circulatory system risks could prevent these people from Everest climbing altogether. As new medical innovation using carbon isotopes will be able to check the tissue age of parts of your heart, lungs etc with a scanner and genetic analysis will be able to make a determination on if you possess the genes to properly produce red blood cells under low pressure at scale, and if your circulatory system can endure the dilation as well -(these are just examples).

    So maybe in the future with most health factors out of the equation, the death rate lowers significantly to just objective dangers and technique mistakes which (using the present data) would get the death rate down to an more palatable .0025% -which is 1 in 400..

    uh…….which still 4x more dangerous than base jumping though.


  • September 3, 2020 at 5:32 pm

    Mark, also, the picture you posted of the North Col, did you take that? if you look close -one of the climbers on the col trail is roped on their way up but in the relatively benign col (excluding high wind or walking on the serac to the left). Is that standard? I don’t recall -did you use a rope there? Without more info, I think if they need roping at the safest spot on the entire climb, how are they going to make it up the giant face, the 3 steps, the rocky traverse, the pyramid, or the descent including the North Col. You’ve been there, maybe you can illuminate an opinion on the quality of climber needed to make the more difficult North side but still needing rope at the easiest spot. I guess it’s possible they were just getting used to the rope system/technique with the guide by roping up at the outset from North Col camp instead of starting with rope in the steeper and more technical terrain further up.

  • September 3, 2020 at 6:25 pm

    Yes, it’s my photo. There are fixed ropes all the way from the North Col up to the summit. They are more for safety than climbing. Although the route is not especially steep, if you fall in pretty much any part above the North Col you could potentially fall a very long way. Having said that, most of us are guilty of hauling on them with our jumars on the steeper sections. Every little thing is exhausting at that altitude.

  • September 3, 2020 at 6:48 pm

    Gotcha, I only see the rope between two people midway and not in the foreground or after -so that’s why I thought it was a personal line between guide and client

  • September 3, 2020 at 6:51 pm

    Mark, did you ever stumble or fall to the point where the rope saved you during the North Col climb and descent?

  • September 3, 2020 at 7:25 pm

    Yes. On the North East Ridge. It’s all described in the book(s) (Seven Steps and also The Chomolungma Diaries, which has a bit more detail about my summit day) 🙂

  • September 3, 2020 at 8:07 pm

    Ha. no excuses. I have several of your books :). My backlog book cue is formidable though. The Everest book is maybe 8 away from reading right now -hopefully get there this month.

    BTW, I climbed in your stomping grounds (well more hiked, scrambled and bouldered) in the Black Cullins on Isle of Skye last year, and if not for the quarantine messing up my travel I’d be there now. Incredible. Not sure I need much more.

  • Pingback:ExWeb’s Links of The Week

  • Pingback:Mark Horrell has 10 facts about Everest success and death rates – Adventure Trend

  • Pingback:She’s Summited Everest 9 Times: Meet Lhakpa Sherpa, A World Record-Holder

Leave a reply

Your email address will not be published, but it will be stored. Please see the privacy statement for more information. Required fields are marked *

Lively discussion is welcome, but if you think your comment might offend, please read the commenting guidelines before posting.

This site uses Akismet to reduce spam. Learn how your comment data is processed.