On 10–11 May 1996, a series of decisions on the Southeast Ridge of Mount Everest culminated in one of the most scrutinised disasters in mountaineering history. Among the dead were two of the most experienced commercial expedition leaders then operating on Everest: Rob Hall, leader of Adventure Consultants, and Scott Fischer, leader of Mountain Madness.
The catastrophe is frequently described as the result of a violent storm. That explanation is true, but incomplete. The storm transformed an increasingly precarious situation into a disaster; it did not create all of the conditions that made the disaster possible.
The accounts of Jon Krakauer, Anatoli Boukreev, Beck Weathers, David Breashears, Ed Viesturs, Jamling Tenzing Norgay, Kenneth Kamler and other participants reveal something more complicated: small compromises accumulated until the climbers had almost no safety margin remaining when the weather deteriorated.
No single decision killed Hall or Fischer. Instead, there was a chain:
congestion → delays → erosion of turnaround discipline → late summits → depleted oxygen and strength → deteriorating weather → separated teams → rescue difficulties → fatalities.
This analysis reconstructs that chain and asks a deliberately counterfactual question:
At what points could different decisions realistically have broken it?
⸻
9 May 1996 — The summit attempt begins
Adventure Consultants and Mountain Madness prepared to leave Camp IV on the South Col at approximately 7,900 metres for their summit attempts.
Between the two commercial teams, together with Sherpas and climbers from other expeditions, a substantial number of people would be moving through the same narrow sections of the Southeast Ridge.
This was the first warning sign.
Everest’s normal climbing difficulties become disproportionately important at extreme altitude. A ten-minute delay at sea level is trivial; above 8,000 metres, standing stationary consumes oxygen, body heat and energy while producing no progress.
The two commercial expeditions therefore needed exceptionally efficient movement.
Instead, they entered the summit day with a system vulnerable to congestion.
Alternative decision
Hall and Fischer could have staggered their summit days or departure times much more substantially.
This would not have eliminated Everest’s objective dangers, but reducing the number of climbers simultaneously negotiating the Balcony, South Summit and Hillary Step would have reduced waiting times and preserved oxygen.
The crucial point is that the eventual disaster did not begin when the storm arrived.
The safety margin was already being consumed during the ascent.
⸻
Late evening, 9 May — The teams commit
At around 10 p.m., the final decision was made to proceed. Contemporary testimony collected by PBS records that conditions appeared sufficiently good to continue. (PBS)
There was nothing inherently reckless about beginning the summit attempt. Everest summit bids routinely begin at night so climbers can reach the summit and descend before evening.
The critical requirement is therefore not merely choosing when to leave.
It is deciding in advance when to stop climbing.
That distinction would become enormously important.
⸻
Early hours, 10 May — Problems with the fixed ropes
As the climbers progressed upward, delays developed at technical sections where fixed ropes had not been prepared as efficiently as anticipated.
Climbers and guides therefore had to spend time establishing ropes while other climbers accumulated behind them.
A bottleneck developed.
This was not merely inconvenient.
Every climber waiting was consuming:
- bottled oxygen,
- body heat,
- physical reserves,
- daylight,
- and the time available for descent.
Boukreev’s account places considerable emphasis on the delays and his involvement in fixing the route. The American Alpine Club’s contemporary review of The Climb similarly notes Boukreev’s role in fixing the Hillary Step after the anticipated preparations had not been completed. (AAC Publications)
Alternative decision
Once significant delays became apparent, the expedition leaders had three increasingly conservative options:
- Impose an earlier turnaround time.
- Turn around clients whose progress indicated that they could not summit sufficiently early.
- Abandon the summit attempt altogether if the cumulative delay became unacceptable.
All would have been psychologically difficult.
But the fundamental principle of high-altitude mountaineering is that the summit is optional; returning is compulsory.
⸻
Sunrise — Congestion at the Balcony
By sunrise, climbers from several expeditions had accumulated around the Balcony.
Beck Weathers later remembered the rising sun almost as a clock: daylight provided warmth, but its movement also demonstrated how quickly the available hours were disappearing.
PBS’s reconstruction records considerable congestion at this point. (PBS)
This was another opportunity to reassess the climb.
The expedition leaders now possessed information they had not possessed when leaving the South Col:
the ascent was progressing more slowly than intended.
Continuing therefore required spending some of the safety reserve planned for the descent.
Alternative decision
Hall and Fischer could have treated the Balcony as a formal decision gate.
Instead of asking:
Can this person still reach the summit?
the question needed to be:
Can this person reach the summit and still possess sufficient oxygen, strength and daylight to descend safely?
Those are fundamentally different tests.
⸻
Morning — Beck Weathers’ eyesight fails
Adventure Consultants client Beck Weathers developed serious visual problems related to previous eye surgery.
He could no longer continue safely.
Hall instructed Weathers to remain where he was and wait for Hall to collect him during the descent.
In isolation this appears reasonable. Weathers was conscious, mobile and relatively close to the route.
But it created another obligation for Hall.
Hall was no longer simply responsible for climbers above him.
He now had a client waiting below him who depended upon his return.
When Lou Kasischke, Stuart Hutchison and John Taske later encountered Weathers during their descent, they offered to take him down. Weathers declined because he had promised Hall that he would wait. (PBS)
Alternative decision
Weathers should arguably have been sent immediately back to Camp IV with a descending climber or Sherpa.
That would have removed one vulnerable person from the upper mountain and one responsibility from Hall.
It illustrates a recurring theme of the disaster:
Decisions that were individually defensible became dangerous when combined with other decisions.
⸻
Late morning — The clock becomes the principal danger
By late morning, it was becoming increasingly apparent that many climbers were behind schedule.
This is where the accounts become particularly revealing.
Lou Kasischke, John Taske and Stuart Hutchison independently recognised the problem and turned around.
They sacrificed Everest’s summit.
They survived.
PBS records their decision around midday: they concluded that they could no longer reach the summit within an acceptable turnaround time. (PBS)
Their decision provides perhaps the most important counterfactual evidence of the entire disaster.
Turning around worked.
They subsequently reached the South Col around the time the storm engulfed it.
⸻
1 p.m. — The critical threshold
A commonly discussed turnaround time for the expeditions was approximately 1–2 p.m.
Yet climbers continued upward.
This is arguably the pivotal decision of 10 May.
The summit is only halfway.
Reaching it at 3 p.m. means beginning the most dangerous part of the expedition — descending exhausted through the Hillary Step, South Summit and Balcony — late in the afternoon.
And the delays had another consequence.
Climbers ascending and descending would now meet at the same technical bottlenecks.
Alternative decision
A non-negotiable turnaround time would probably have prevented much of what followed.
Not:
“We’ll turn around at 1 p.m. unless we’re nearly there.”
But:
At the predetermined time, everyone who has not summited descends.
No exceptions for proximity.
No exceptions for previous failed attempts.
No exceptions because a client has paid tens of thousands of dollars.
No exceptions because the weather still looks good.
The rule exists precisely because judgement deteriorates through exhaustion, hypoxia and summit fixation.
⸻
1:12 p.m. — Krakauer reaches the summit
Jon Krakauer reached the summit at approximately 1:12 p.m.
He did not remain there long.
He began descending while numerous climbers were still ascending.
Outside’s later reconstruction records him noticing clouds spreading across Lhotse during his descent. (Outside Online)
This is significant because the expedition was now approaching the point where early summiters were descending through climbers still attempting to reach the top.
The route was beginning to work against itself.
⸻
Approximately 2 p.m. — The summit becomes crowded
More climbers reached the summit around 2 p.m.
Contemporary reporting noted immediately that these were very late Everest summits. NOVA’s dispatch on 10 May specifically expressed concern that the climbers had summited at around 2 p.m. and still had to reach the South Col. (PBS)
Michael Groom left the summit at approximately 2:30–3 p.m.
On his descent he encountered Adventure Consultants client Doug Hansen still climbing, approximately 70 metres below the summit. (PBS)
This was an extraordinarily consequential moment.
Hansen had attempted Everest with Hall the previous year but failed to reach the summit.
Hall remained with him.
The human motivation is understandable.
The mountaineering consequences were catastrophic.
⸻
The Doug Hansen decision
Hall’s relationship with Hansen lies at the heart of the tragedy.
Hansen was close to achieving something for which he had invested enormous money, effort and emotion.
Turning him around within sight of the summit would have been devastating.
But Hall was not merely Hansen’s climbing companion.
He was his guide.
The guide’s responsibility sometimes requires making precisely the decision the client does not want made.
By allowing the ascent to continue, Hall effectively exchanged descent margin for summit success.
Alternative decision
Hall should have turned Hansen around when the agreed turnaround threshold was exceeded.
Had that occurred at roughly 1–2 p.m., Hall and Hansen could have been descending before the worst deterioration.
We cannot know that both would have survived.
But their probability of survival would have been dramatically greater.
⸻
Approximately 3 p.m. — Scott Fischer reaches the summit
Fischer was also extraordinarily late.
He had experienced a difficult expedition and was reportedly fatigued before summit day.
His responsibilities as expedition leader had repeatedly required him to move up and down the mountain.
By the time he approached the summit, he was considerably depleted.
Mountain Madness guide Neal Beidleman eventually ordered clients downward rather than waiting indefinitely for Fischer.
Outside’s reconstruction places that decision at approximately 3:12 p.m.; Fischer was still ascending. (Outside Online)
Fischer subsequently reached the summit at roughly 3:30 p.m.
For a commercial expedition leader, this was dangerously late.
Alternative decision
Fischer had perhaps the clearest opportunity of anyone to save himself:
turn around.
His clients had already summited.
His professional objective had effectively been achieved.
Continuing to the summit added little to their safety or success.
But it consumed the physical reserves Fischer desperately needed for descent.
⸻
The critical difference between Hall and Fischer
Although their deaths are frequently grouped together, the circumstances were different.
Hall stayed high primarily because he remained with a client in trouble.
Fischer became the person in trouble.
Hall’s fatal problem was therefore partly one of guide responsibility.
Fischer’s was partly one of self-management and accumulated exhaustion.
Both, however, shared the same underlying problem:
they were far too high, far too late.
⸻
After 3 p.m. — The mountain begins closing behind them
Descending climbers encountered ascending climbers at the Hillary Step.
Another bottleneck developed.
Charlotte Fox later recalled the concern that reaching the summit represented only half the climb and that everyone still had to get home. (PBS)
By now several safety margins had disappeared simultaneously:
daylight was declining;
oxygen supplies were diminishing;
climbers were exhausted;
the teams had become dispersed;
guides were separated from clients;
and the weather was deteriorating.
Any one of these could potentially have been managed.
Together they constituted a system approaching failure.
⸻
Approximately 3:20 p.m. — Hall and Hansen descend
Hall finally began descending with Hansen.
They were extraordinarily late.
Hall was now attempting to guide an exhausted client down technically difficult terrain while both were approaching the limits of their physiological reserves.
The earlier decision not to enforce the turnaround time had now become almost impossible to reverse.
This is an important feature of disasters.
A bad decision made early often removes good options later.
Hall could no longer simply “turn around.”
He had turned around.
The problem was that he had done so several hours too late.
⸻
Late afternoon — The storm arrives
The weather deteriorated rapidly.
Those who had turned around earlier were reaching the South Col.
John Taske recalled seeing the storm engulf Camp IV as he approached it. Within moments visibility collapsed into whiteout conditions. (PBS)
This is crucial when considering whether “the storm killed them.”
The storm unquestionably caused the final crisis.
But Kasischke, Hutchison and Taske encountered essentially the same storm.
They were simply lower when it arrived.
The difference was created hours earlier.
⸻
4 p.m. — Hall and Hansen remain high
At approximately 4 p.m. Hall radioed that he and Hansen remained near the Hillary Step.
They were now roughly three hours beyond the nominal turnaround threshold. (PBS)
The expedition had moved from prevention to rescue.
And rescue above 8,500 metres is extraordinarily difficult.
A rescuer cannot simply carry another adult down Everest.
Every person sent upward is themselves placed at risk.
⸻
5:14–5:35 p.m. — Hall is no longer making progress
Radio communications revealed the seriousness of the situation.
At 5:14 p.m. Hall indicated that he was moving.
At 5:35 p.m. he was apparently still in essentially the same location.
Guy Cotter, communicating with Hall, recognised what this meant.
Hall could not move Hansen effectively.
Cotter eventually raised the almost unbearable option:
Hall might have to leave Hansen and save himself.
Cotter later explained the brutal logic: a rescuer who cannot rescue the casualty must avoid becoming another casualty. (PBS)
Alternative decision
At this point the choices were terrible.
Hall might have abandoned Hansen and descended alone.
Whether he could still have reached safety is unknowable.
But the decision illustrates how drastically the available options had deteriorated.
At 1 p.m., Hall’s choice had been:
summit or descend.
At 5:30 p.m., it had become:
remain with a dying client or abandon him in an attempt to survive.
That transformation is the essence of the disaster.
⸻
Approximately 6 p.m. — Fischer collapses
Fischer’s deterioration followed a different course.
Makalu Gau encountered him below the South Summit area. Fischer was no longer walking normally and was reduced to sitting and sliding down the slope.
Eventually Fischer became effectively immobile around 8,300 metres/27,100 feet.
Gau was also incapacitated nearby.
The storm engulfed them.
They survived the night exposed on the mountain, but Fischer’s condition deteriorated beyond recovery. (PBS)
Could Fischer have been rescued?
Possibly — but only earlier.
Once an incapacitated man weighing perhaps 80–90 kg is stranded above 8,000 metres during a storm, rescue options become extraordinarily limited.
The effective rescue decision therefore wasn’t made at 6 p.m.
It had been available hours earlier when Fischer still possessed the ability to descend under his own power.
⸻
Evening — The South Col becomes a second disaster
Meanwhile, numerous descending climbers became lost only a few hundred metres from Camp IV.
Michael Groom, Neal Beidleman, Beck Weathers, Yasuko Namba and others became trapped in the storm.
They could communicate with camp but could not see it.
At times they were perhaps only 100–200 metres from safety. (PBS)
This demonstrates how completely the weather had transformed the mountain.
An exhausted climber could literally be within walking distance of a tent and still be unable to find it.
⸻
Boukreev’s controversial descent
One of the most debated decisions concerned Anatoli Boukreev.
Boukreev had descended ahead of many Mountain Madness clients and reached Camp IV.
Krakauer later criticised this.
Boukreev and DeWalt vigorously defended it in The Climb.
Their argument was essentially that Boukreev returned to camp so that he could recover, hydrate and be available if assistance became necessary.
And assistance became necessary.
Boukreev repeatedly left the tents during the storm, located stranded climbers and brought several back to camp.
The controversy therefore resists a simple judgement.
A guide being separated from clients during descent appears undesirable.
But Boukreev’s presence at Camp IV with sufficient reserves to conduct rescues unquestionably saved lives.
The American Alpine Club’s review of The Climb specifically notes this rationale. (AAC Publications)
The better lesson may therefore be organisational rather than personal:
the expedition needed a predetermined rescue-capable guide at Camp IV rather than relying upon an improvised arrangement.
⸻
Night of 10–11 May — Hall remains above the South Summit
Hall survived the night high on Everest.
Hansen apparently died sometime during the night.
Guide Andy Harris attempted to assist Hall and subsequently disappeared.
Hall eventually obtained oxygen and remained in radio contact.
A rescue attempt was organised, but reaching him in those conditions proved impossible.
By this stage the earlier decisions had become irreversible.
Hall was effectively stranded close to the summit of Everest.
⸻
Morning, 11 May — The cruel contrast
Morning revealed an extraordinary situation.
Some climbers assumed dead were still alive.
Beck Weathers regained consciousness after spending the night exposed and somehow walked into Camp IV.
Makalu Gau was also alive near Fischer.
But Hall remained extremely high.
The rescuers faced the fundamental limitation of high-altitude rescue:
saving somebody at 7,900 metres is extraordinarily difficult; saving somebody near 8,700 metres may be physically impossible.
⸻
11 May — Fischer is found
Sherpas reached Fischer and Gau.
Gau could still be rescued.
Fischer could not.
Contemporary reports describe attempts at resuscitation, but he was beyond recovery. (PBS)
The contrast between Gau and Fischer again illustrates how narrow the margins were.
Two men could spend the same night close together and experience different outcomes because of relatively small differences in exhaustion, physiology, oxygenation and exposure.
⸻
Hall’s final hours
Hall remained alive high on the mountain.
Attempts to reach him failed.
He communicated by radio with Base Camp and ultimately with his pregnant wife, Jan Arnold, in New Zealand.
He died high on the Southeast Ridge.
His body remained where he died.
The irony is particularly poignant because Hall had been regarded as one of the most cautious and accomplished commercial Everest guides of his generation. The American Alpine Club’s obituary describes a climber who deliberately used supplementary oxygen above 8,000 metres and placed great emphasis on retaining sufficient reserves to help weaker climbers. (AAC Publications)
On 10 May 1996, the system of safeguards he normally valued progressively failed.
⸻
Could the disaster have been prevented?
With hindsight it is tempting to identify one mistake.
There wasn’t one.
There was a cascade.
- Too many climbers occupied the same route
This created delays at precisely the places where movement needed to be efficient.
Alternative: stagger summit attempts and coordinate departures between expeditions.
- Fixed-rope preparation was incomplete
Time and oxygen were consumed establishing protection during summit day.
Alternative: verify critical ropes before committing clients above Camp IV.
- Delays were not translated into revised turnaround times
The teams continued operating according to summit ambitions despite losing substantial time.
Alternative: every hour lost ascending should reduce the allowable summit deadline.
- Turnaround times became flexible
This was probably the most consequential failure.
Alternative: make turnaround time absolute and empower every guide to enforce it.
- Hansen continued too late
Hall’s understandable desire to help Hansen achieve his summit dramatically reduced both men’s survival margin.
Alternative: turn Hansen around regardless of proximity to the summit.
- Fischer continued despite exhaustion and lateness
Fischer had already expended enormous energy during the expedition.
Alternative: Fischer turns around and delegates summit supervision to Beidleman and Boukreev.
- Weathers remained waiting
A vulnerable climber remained high unnecessarily.
Alternative: attach him immediately to the first competent descending group.
- Guides and clients became dispersed
When the emergency developed, guides were scattered across the mountain.
Alternative: predetermined guide positions — lead, middle, sweep and Camp IV reserve.
- The expeditions lacked sufficient contingency for simultaneous casualties
Commercial guiding assumes guides assist clients.
On 10 May, several guides and expedition leaders themselves became casualties.
Alternative: designate personnel specifically for emergency support rather than committing virtually everyone to summit operations.
⸻
The decision that probably mattered most
If one intervention has to be selected, it is surprisingly simple:
At 1–2 p.m., everybody who had not reached the summit turns around.
That single rule potentially changes almost everything.
Hansen descends with Hall.
Fischer abandons his own summit attempt.
Hall does not become stranded waiting for Hansen.
Fischer retains substantially more physical reserve.
Descending climbers pass the Hillary Step earlier.
Oxygen reserves are greater.
The teams are lower when the storm arrives.
Weathers is collected earlier.
Guides and clients reach the South Col before darkness.
It cannot be claimed that everyone would certainly have survived.
Everest never permits that certainty.
But the chain leading to the deaths of Hall and Fischer would have been broken at several points simultaneously.
⸻
Conclusion — The storm was the trigger, not the whole cause
The tragedy of Everest in May 1996 should not be reduced either to “a freak storm killed the climbers” or to “reckless guides ignored obvious danger.”
Both explanations are too simple.
Hall and Fischer were accomplished mountaineers operating in an environment where judgement is degraded by hypoxia, exhaustion, cold, commercial responsibility and enormous psychological investment.
They did not knowingly choose death.
Instead, they repeatedly accepted slightly smaller margins of safety.
A delayed rope.
Another twenty minutes.
A client who was nearly there.
A summit that remained tantalisingly close.
Good weather that appeared likely to last.
Enough oxygen for just a little longer.
Each decision was understandable in isolation.
Collectively they were disastrous.
The most important evidence comes from the people who turned around.
Lou Kasischke, Stuart Hutchison and John Taske confronted exactly the psychological dilemma faced by the summit climbers. Everest was above them. Months of preparation and enormous expense lay behind them.
They nevertheless recognised that the clock mattered more than the summit.
They descended.
When the storm arrived, they were approaching Camp IV rather than struggling near 8,700 metres. (PBS)
That provides perhaps the clearest lesson of 1996.
The disaster was not inevitable when the climbers left the South Col. Nor was it necessarily inevitable when the ropes caused delays, when the Balcony became congested, or even when the first climbers reached the summit. At each stage there remained an opportunity to break the chain.
But each opportunity became more expensive.
At noon, abandoning the summit meant disappointment.
At 2 p.m., it meant accepting failure within sight of the goal.
At 5 p.m., Hall faced abandoning Doug Hansen.
By nightfall, there were virtually no good decisions left.
That is why the most enduring lesson of Everest 1996 is not simply about storms, oxygen or commercial mountaineering.
It is about decision-making while choices still exist.
A safe turnaround can feel unnecessarily conservative precisely because the catastrophe it prevents never happens.
Hall and Fischer were caught by a severe storm. But the storm became lethal because, by the time it arrived, time, oxygen, daylight, physical reserves and organisational resilience had already been spent pursuing the summit.
Everest did not leave them no choice.
The tragedy was that the safest choices had been available hours earlier.
