Ski Deaths in February 2026: A Tragic Overview
The snowpack is an honest record-keeper. It has no editorial interest in what it stores. Each storm leaves a layer, each dry spell leaves a gap, each freeze-thaw cycle leaves its chemical signature in the crystals. Snow scientists read these strata the way a physician reads an X-ray – methodically, without sentiment, looking for the thing that should not be there. In the Sierra Nevada, in January 2026, what should not have been there was everywhere: a persistent layer of weak, faceted snow, sugar-like in structure and useless as a foundation, formed during a long dry spell and buried when the storms finally arrived. The Sierra Avalanche Center tracked it. The Utah Avalanche Center tracked a nearly identical formation in the Wasatch. They published their findings in daily forecasts. They named the problem. They explained, in plain language, what it meant.
What it meant was this: when the next major storm cycle arrived, the new snow would have nowhere to bond. It would sit on top of the old, heavy and unattached, waiting for a reason to move.
The storms came in mid-February. In some locations they delivered three to four inches of snow per hour. The snowpack, which had been keeping its record faithfully all winter, delivered its verdict.
What followed was the deadliest month in American skiing in recent memory. Not because mountains became more dangerous. Because the forces the industry has been managing around – and in some cases managing away from public view – ran out of runway all at once.
The Record
By the final week of February 2026, at least sixteen skiers and snowboarders had died across the American West. Nine of them in a single avalanche in the Sierra Nevada, the deadliest slide in the United States in 45 years. Three at a single resort in the Lake Tahoe basin, over a span of two weeks. Two more at a different resort on the same lake, in the same morning, in unrelated incidents. An 11-year-old girl from Massachusetts on a family ski vacation in Utah. A father killed in a backcountry slide while his son watched and dug.
These deaths occurred across California, Utah, and Idaho. They involved resort skiers, guided backcountry clients, sidecountry adventurers, and a child who crossed a rope line into uncontrolled terrain. They were different in almost every circumstance of their occurrence. They shared one thing: the same anomalous snowpack, formed by the same meteorological sequence, had loaded the same trap across an entire region of the American West simultaneously.
The industry’s instinct, when a month like this lands, is to treat each death as its own isolated event. The collision on the advanced trail. The medical emergency. The solo skier who did not return to his car. The guided group caught in an historic avalanche. The child who crossed the boundary. Each one, in isolation, fits neatly into the framework the industry has spent decades constructing: skiing involves inherent risk, participants accept that risk, and individual decisions drive individual outcomes. That framework is not wrong. It is, however, incomplete.
The Dissolving Line
Somewhere in the past decade, the line between resort skiing and backcountry terrain became more of a suggestion than a boundary. The industry did not draw the line in the wrong place. It drew it correctly, clearly, with rope and signage and patrol infrastructure on one side and wilderness on the other. What changed was the culture – and to some degree the commercial apparatus – around the line itself.
The backcountry boom that accelerated after 2020 brought a surge of new participants into terrain that does not grade on a curve. Touring gear sales hit records. Avalanche safety courses filled months in advance. Social media made the backcountry legible and aspirational in ways it had never been before – footage of untracked powder fields, hut-to-hut narratives, the aesthetic of self-propelled wilderness earned by people who looked competent and alive. What social media did not convey, with any consistency, was the texture of the snowpack beneath those pristine runs, or the distinction between a guided client and a mountain professional, or the specific and non-negotiable difference between terrain managed by ski patrol and terrain managed by gravity alone.
The commercial infrastructure followed the boom. Guided backcountry hut packages, priced at $1,795 for a three-to-four-day stay, marketed to experienced resort skiers looking for the next level of adventure. Sidecountry gates at resorts, which make it trivially easy to exit the boundary into uncontrolled terrain, often with no more friction than a posted sign and a rope. Terrain expansion at resorts pushing further into glades and tree runs that back against unpatrolled slopes. The industry, rationally, served the market. The market wanted access. Access was sold.
What was not sold, with anything like equivalent force, was the education required to use that access safely. A family from Massachusetts, on February school vacation, crosses a rope line at a ski resort. On one side of that line is a groomed, patrolled, avalanche-controlled mountain. On the other side is backcountry terrain, subject to no control work whatsoever, where the Utah Avalanche Center had issued a full warning that morning. The family carries no avalanche rescue equipment. They do not know – perhaps because nobody told them in terms that registered – that the rope line is not a traffic cone but a cliff edge. The 11-year-old girl triggers a slide in Mary Main Bowl. The avalanche buries her four feet deep. The family uses a phone app to find her. Ski Patrol performs CPR for thirty minutes. A rescue helicopter cannot land due to weather. She does not survive.
The Utah Avalanche Center’s final report on that incident included a line that deserves to be read as policy: “Anyone leaving a ski area boundary must treat the terrain they enter as backcountry.” Utah, the report noted, historically leads the nation in avalanche fatalities involving people who leave ski area boundaries. That statistic did not emerge from nowhere. It emerged from a persistent, structural failure to communicate to resort visitors – who arrive in large numbers, many of them tourists, many of them unfamiliar with the terrain – what the rope line actually means.
The industry is not oblivious to this. Resorts post signage. Patrol talks to sidecountry users. The Skier’s Responsibility Code, overhauled in 2022, addresses boundary compliance. But signage and code updates are passive instruments in a culture that has been actively marketed into believing that adventure is adjacent – just past the rope, just around the ridge, just one level up from what you already know how to do. The industry profits from that belief and educates against it with roughly equal commitment. The result is a boundary that is clear on paper and porous in practice.
The Information That Existed
One of the harder facts about February 2026 is that the information needed to prevent most of what happened was available, accurate, and free. The Sierra Avalanche Center issued a Level 4 HIGH danger warning – one step below the maximum – at five in the morning on the day of the Castle Peak avalanche. It said, explicitly, that large avalanches were expected and that travel in, near, or below avalanche terrain was inadvisable. Interstate 80 through the Sierra was closed. Several nearby resorts had suspended operations entirely. The guiding company at the center of the Castle Peak tragedy had themselves posted a video to social media the previous day observing a “particularly weak layer” in the snowpack and urging extra caution.
The Utah Avalanche Center had issued its own warning before the avalanche that buried an 11-year-old girl at Brighton’s boundary. The father who died in the Wasatch died during what the Utah Avalanche Center described as the range’s most volatile avalanche cycle of the season.
All of this information was broadcast. All of it was accurate. And yet people went out.
The temptation, here, is to conclude that the problem is individual decision-making – that people hear warnings and choose to ignore them, and that the consequences of that choice are their own. That conclusion has the virtue of being partially true and the liability of being profoundly insufficient. The Castle Peak group included certified guides who held American Mountain Guides Association credentials and avalanche instructor certifications. They had read the forecast. They were on the last day of a three-day trip and were trying to exit early, reducing their time on exposed terrain. They carried every piece of safety equipment the industry recommends. None of it was enough, when a football field-sized slab of snow classified between a D2 and D3 on the destructive size scale released on terrain that veteran patrollers said they had never seen slide before.
The problem is not that warnings go unread. The problem is that the gap between “I read the warning” and “I understand what this terrain will do today” is wider than any single forecast can bridge for any single person in the field. Avalanche education has improved markedly over the past two decades. Colorado made it to the end of February 2026 without a single avalanche fatality – a feat not accomplished in 22 winters – while California and Utah were absorbing the losses described above. It was not that Colorado’s mountains were safer. Colorado has spent two decades building the gap-closing infrastructure that other states have not. But education is not evenly distributed. It does not reach the family crossing a rope line on February vacation. It does not fully reach the guided client who has handed over their terrain decisions to a professional and trusts that the professional has resolved the question.
What the Industry Does Not Count
After five people died at ski resorts owned by a single corporation in the Lake Tahoe basin in February, CalMatters asked a question that should have a simple answer: Is California’s ski season deadlier than usual? The answer, from the state’s official data infrastructure, was that nobody knows. California does not monitor ski injuries or deaths at resorts. It does not require resorts to report fatal incidents. It has no threshold for injury rates that triggers investigation or regulatory intervention. CalMatters contacted more than two dozen California ski resorts. Not one provided accident, injury, or fatality data. A public records request to the U.S. Forest Service for five years of incident reports from resorts operating on federal land produced an estimate: six months to process, because the resorts themselves must first review the records to determine what they consider proprietary.
John Rice, the president of Ski California, offered a response to the February deaths that is worth examining carefully. “There’s been no indication that there are more injuries this year than previous years,” he said, “just more media coverage around serious ones.” He may be right. The industry may be performing exactly as it historically has, and February 2026 may be an outlier produced by extraordinary meteorological conditions rather than any systemic deterioration in safety. But Rice cannot prove that claim, and neither can anyone else, because the data does not exist in any form accessible to the public, to researchers, to lawmakers, or to journalists.
This is not an accident. It is a posture, refined over decades of legislative resistance. A 2010 California bill authored by Assemblymember Dave Jones would have required ski resorts to publish monthly fatality reports – cause, location, age, where medical care was provided – and maintain publicly available safety plans. Governor Arnold Schwarzenegger vetoed it. A subsequent two-year effort to mandate monthly accident reporting to the California Department of Public Health died in the legislature before it reached a governor’s desk. Ski California opposed it. The industry’s argument, stated plainly, is that numbers without context are easily misinterpreted; that terrain, weather, and visitor behavior vary so widely between resorts that raw data is misleading; and that confidentiality laws govern incident information handled by ski patrol. All of that is arguable. None of it explains why the industry that manages the risk cannot be required to count the outcomes.
Dan Gregorie, a retired physician whose daughter Jessica died at Alpine Meadows ski resort in 2006 after sliding off an unguarded traverse and over a cliff, spent over a decade personally funding lobbying efforts to change this. He hired a lobbyist. He fought in California, Colorado, and Maine. He founded the SnowSport Safety Foundation. He lost his daughter. Two years later he lost his wife to cancer. He kept going. The SnowSport Safety Foundation is no longer active. Gregorie has given up on legislation. He still tells people, when they mention a ski trip, “Do you know what you’re going into?” The answer, structurally, is that they cannot – because the industry that could tell them has chosen not to.
Former Assemblymember Jones made a point the industry has never satisfactorily rebutted: the argument for data transparency is not primarily that individual consumers would change their behavior. It is that publicly available safety data allows consumer advocacy organizations, researchers, and safety experts to identify patterns, compare resort performance, and press for systemic changes. Competition around safety only functions if safety is visible. Food safety, aviation safety, highway safety – every industry that handles large numbers of people in high-risk environments publishes some version of incident data, and every one of those industries is measurably safer for it. Skiing is not structurally different from these industries in the risks it creates. It is different in the latitude it has been granted to manage those risks privately, with the outcomes locked behind proprietary claims and legal waiver doctrine.
The Greater Problem
February 2026 did not create the conditions that killed sixteen skiers and snowboarders. It revealed them. The forces at work have been building for years, and they are not retreating.
The backcountry boom is structural, not cyclical. Equipment manufacturers reported record touring gear sales in consecutive seasons through the early 2020s. Backcountry visits have been growing at rates that double every four to five years in some regions. The population of people moving through avalanche terrain has expanded far faster than the population that has received meaningful avalanche education. Those two curves have been diverging for years, and nothing on the horizon closes the gap quickly.
Climate volatility is complicating the snowpack in ways that make February-style scenarios more probable, not less. The dry-spell-and-violent-storm pattern that created the persistent weak layers across California and Utah in 2026 is consistent with the kind of snowpack instability that researchers associate with a warming climate – longer dry periods, more intense precipitation events, more freeze-thaw cycling at elevations that historically stayed cold. The Sierra Nevada and the Wasatch are not going to become more forgiving. The snowpack is going to keep its record honestly. What changes is the character of what it records.
Resort crowding is pushing skiers and riders toward the margins in ways the industry has not fully grappled with. When the groomed terrain at a major resort is dense with bodies on a powder day, the sidecountry gate becomes attractive, and the tree run at the edge of the boundary becomes the path of least resistance toward untracked snow. This is not a pathology – it is a rational response to the product the industry sells at the price it charges. Packed powder days at premium-priced mega-resorts are a natural consequence of the Epic Pass model that aggregates skier visits at a small number of dominant destinations. The terrain has not expanded proportionally. The margins are where the overflow goes.
The guided backcountry sector is growing into a regulatory vacuum. Large groups of people in expert-level terrain, under the supervision of guides whose qualifications vary and whose go/no-go decisions exist in a framework with no external accountability. The Castle Peak investigation will eventually determine whether Blackbird Mountain Guides made a defensible decision or a negligent one on February 17. What the investigation cannot determine is how many similar decisions are made every day across the American West, against similar forecasts, in similar terrain, with no regulatory framework to define what defensible looks like before the outcome reveals it.
There is no single intervention that prevents another February. But there is a set of interventions, each well within the industry’s capacity, that would make the next convergence of bad snowpack and expanding backcountry participation significantly less lethal. Mandatory incident reporting – not as punishment but as data – would create a baseline against which trends could be measured and patterns could be identified. Investment in avalanche education infrastructure, particularly in states and regions that have not built Colorado’s culture of information application, would close the gap between warnings and decisions. Clearer, more forceful, more redundant communication at resort boundaries – not just signage but active patrol engagement, mandatory awareness content at ticket purchase, standardized language about what lies beyond the rope – would reach the family from Massachusetts before they cross the line rather than after. And a regulatory framework for guided backcountry operations, even a minimal one establishing baseline qualification standards and documentation requirements for go/no-go decisions in high-danger conditions, would introduce at least some accountability into a space that currently operates largely on professional honor.
None of these things will happen quickly. The industry’s relationship with regulation has been adversarial for decades, and the legal framework that shields resorts from liability has been remarkably durable. The skier’s waiver – that piece of paper everyone signs and nobody reads – has survived every legislative challenge that February’s dead could not.
The Snowpack Keeps Its Record
The Sierra Nevada is recording this season too. The Wasatch is recording this season too. Somewhere in the current snowpack, there are layers that will matter in a storm cycle no one has predicted yet – weak facets buried under recent snow, interfaces that have not bonded, terrain features that have never produced a slide within living memory and are building toward one. The avalanche centers are tracking all of it. They will publish their findings. They will use plain language to describe the problem. Some people will read it. Some of those people will make better decisions because of it.
The industry will record almost none of what happens on its slopes this season in any form accessible to the public. The outcomes of that choice – how many injuries, how many deaths, what patterns exist across what terrain types, what interventions might have mattered – will remain in the same locked room they have always occupied, accessible only through the slow machinery of a public records request or the brute effort of a coroner’s database survey or the persistence of a grieving father who spent a decade of his own money trying to change a system that did not want to be changed.
February 2026 was not the first month that revealed this gap. It was the loudest. Sixteen skiing and snowboarding deaths in a single month is a number that forces a reckoning, or should. The deadliest U.S. avalanche in 45 years is a fact that does not belong in a single news cycle and should not be permitted to become one. The question the industry has to answer – has always had to answer, but has been able to defer because no single month made the avoidance fully visible – is whether it intends to be a partner in the infrastructure of safety or a beneficiary of its absence.
The snowpack does not have a preference. It will record whatever the winter gives it, faithfully and without editorial interest, in layers as legible as any safety report the industry has declined to publish. The avalanche centers will keep issuing their warnings. The rope lines will keep marking their boundaries. The question is whether the industry decides that the gap between those honest signals and the people who need them is its problem to solve – before the next February makes the answer for it.
