TL;DR: California regulators cited training and assessment for airbags, not ownership, after three fatal Sierra avalanches.

California's Division of Occupational Safety and Health issued citations in three fatal Sierra Nevada avalanche cases between February 2025 and August 2026, and avalanche airbags appear in the citations for all three, according to Teton Gravity Research.

The cases involved two different employer types, a large resort ski patrol and a small guide service. The recurring allegation is narrower than a shortage of equipment. In two Mammoth Mountain patrol deaths, the agency alleged failure to ensure workers could deploy airbags effectively. In the Castle Peak guiding deaths, the agency alleged failure to assess whether guides required airbags and AvaLungs for travel in avalanche terrain.

All three employers have appealed. Mammoth Mountain says it has corrected the violations cited after the second patrol death, according to Teton Gravity Research. Blackbird Mountain Guides has filed a formal appeal, according to the same report and a separate account in VeloMag.

Two patrol deaths at Mammoth Mountain

Feb. 14, 2025: Claire Murphy in Avalanche Chutes

Claire Murphy, a 25-year-old patroller, was killed in an avalanche in the Avalanche Chutes area while working to open terrain on Feb. 14, 2025, according to Teton Gravity Research.

California's Division of Occupational Safety and Health issued three citations totaling $26,810 after that death. Among them was a finding that the resort did not ensure patrollers could effectively deploy avalanche airbags. Mammoth appealed, according to Teton Gravity Research.

The accident location name and the work assignment matter for gear planning. The worker was opening terrain, not recreating. The citation did not allege that airbags were absent. It alleged that effective deployment was not ensured.

Dec. 26, 2025: Cole Murphy on Lincoln Mountain

Cole Murphy, a 30-year-old patroller with no relation to Claire Murphy, was caught during mitigation work on Lincoln Mountain on Dec. 26, 2025, according to Teton Gravity Research.

He was buried roughly 15 minutes and later died. California's Division of Occupational Safety and Health issued three citations totaling $126,000 after that death.

The largest citation, $90,000 and classified as willful, found that the resort did not ensure employees were properly instructed in airbag deployment. The other two citations, $18,000 each, concerned communications during mitigation and confirming personnel were clear of the path before explosives were used. Mammoth says it has corrected the violations and is appealing, according to Teton Gravity Research.

The burial interval is directly relevant to a bug-out bag built for winter travel. Roughly 15 minutes under snow leaves little margin for improvised rescue. A device that requires a handle pull in the first seconds of a slide cannot help if the pull has never been rehearsed.

Nine deaths on a Castle Peak hut trip

Feb. 17, 2026: Blackbird Mountain Guides group

Nine people died on a Blackbird Mountain Guides hut trip on Feb. 17, 2026, according to Teton Gravity Research. The dead included guides Andrew Alissandratos, Nicole Choo and Michael Henry, along with six clients.

The group had spent two nights at the Frog Lake ski huts before returning toward the trailhead on the third day, according to The Guardian Mirror. The huts were described as upscale cabins equipped with heaters, wall sockets and bathrooms.

The emergency call came just before noon on Feb. 17, according to The Guardian Mirror. The slide occurred on Castle Peak, a 9,110-foot mountain north of Donner Summit in the Lake Tahoe area.

California's Division of Occupational Safety and Health issued seven citations totaling $151,300 after the Castle Peak deaths, according to Teton Gravity Research. One serious violation, carrying $16,200, alleged Blackbird failed to assess whether guides required protective equipment for travel in avalanche terrain. That citation named avalanche airbags and AvaLungs specifically. Blackbird has filed a formal appeal.

A separate account in VeloMag described the same enforcement action as a fine of over $150,000 and cited failure to keep guides out of avalanche terrain when danger was rated high, attributing that detail to kunr.org. The two descriptions address different alleged violations from the same enforcement package. One concerns equipment assessment. The other concerns terrain selection during high danger.

Rescue conditions limited what gear could do

The Castle Peak rescue shows why a loadout must support sheltering and signaling after a slide, not only flotation during a slide. Six people survived and were rescued, according to multiple reports.

The University of California, Berkeley's Central Sierra Snow Lab near Donner Pass reported 28 inches of snow that day, with another 3 feet expected in the next two days, according to The Guardian Mirror. The risk of another avalanche during rescue was described as high.

The sheriff's office marshaled 46 first responders after the emergency call, according to The Guardian Mirror. Rescue teams traveled by Sno-Cat, a truck-sized vehicle with tank-like tracks for polar conditions, then skied the final distance.

Rescuers reached a point about 2 miles from where beacons indicated survivors were sheltering, then skied the remaining 2 miles of mountainous terrain until they reached the six survivors at about 5:30 p.m., according to The Guardian Mirror. In the hours since the avalanche, survivors had built a shelter from tarps and uncovered three bodies from their group. First responders found five additional bodies at that stage.

All nine bodies were brought off the mountain on Saturday with helicopters and Sno-Cats, according to The Guardian Mirror. That timeline underscores a hard winter-evacuation fact: even with beacons, tarps, tracked vehicles and aircraft, recovery extended over days because of storm loading and continued avalanche danger.

What was actually cited: instruction and assessment

None of the citations described by Teton Gravity Research said the operators lacked airbags. That distinction shapes the equipment lesson.

In the Mammoth cases, the allegation concerned instruction, including whether patrollers had been trained to deploy the device. The wording cited included failure to ensure patrollers could effectively deploy avalanche airbags after the February 2025 death, and failure to ensure employees were properly instructed in airbag deployment after the December 2025 death.

In the Blackbird case, the allegation concerned assessment, including whether the company had evaluated as a documented hazard-assessment step whether guides needed airbags for the terrain and conditions encountered at work. The citation named avalanche airbags and AvaLungs specifically and carried $16,200.

Teton Gravity Research reported that California's Division of Occupational Safety and Health is treating an avalanche airbag like other personal protective equipment on a California worksite. Under that framework, an employer assesses whether the job requires the equipment, provides it when required, and trains the worker for use under real conditions. Possession alone does not satisfy that sequence.

Why deployment failure dominates the data

The most-cited peer-reviewed work on airbag effectiveness is a 2014 study in Resuscitation by Pascal Haegeli and colleagues, which examined 424 seriously involved avalanche victims worldwide, according to Teton Gravity Research.

A deployed airbag cut mortality roughly in half, from 22% to 11%. Once non-deployments were included, the benefit fell to a 41% reduction because 20% of people in the dataset did not or could not pull the trigger.

Later analysis of non-inflation cases attributed roughly 60% to user deployment failure, with device failure and maintenance errors accounting for the remainder, according to Teton Gravity Research. Non-trigger rates were about 5% among avalanche professionals versus about 14% among recreational users.

That gap points to rehearsal rather than purchase. Professionals failed at roughly one-third the rate of recreationists. Teton Gravity Research described deployment as a trained motor skill performed in the worst three seconds of a burial event, and reported that manufacturers sell practice cartridges and refill services so the trigger can be pulled for real, more than once, before it counts.

What this means for a winter bug-out bag

For Prepper Packs readers, the Sierra cases argue for a different packing standard: do not count an undeployed airbag as rescue capability.

A winter evacuation loadout that includes an airbag pack should also include the training record behind it. The cited Mammoth allegations focused on whether workers could deploy under operational stress. The cited Blackbird allegation focused on whether leaders had assessed the need for airbags and AvaLungs before entering avalanche terrain. Both questions can be applied to a personal kit before a storm trip.

Practical steps supported by the reporting include verifying handle access with gloves and a loaded pack, completing a live practice deployment with a practice cartridge, arranging refills or recharge after practice, checking leg-loop or harness connection, and carrying beacon, probe, shovel, tarp shelter and communications alongside the airbag. Teton Gravity Research emphasized that no operator was cited merely for lacking airbags, while Peak Experience NZ separately advises that an airbag never replaces conservative decision-making, beacon, shovel, probe, trained partners or rescue practice.

The Castle Peak death toll also counsels against treating any single device as an entry ticket to high-danger terrain. Nine dead, six survivors, 28 inches of new snow in one day, 3 more feet forecast, and a multiday helicopter and Sno-Cat recovery show the scale of exposure when travel continues through a major loading event, according to The Guardian Mirror. An airbag may improve odds during motion through debris, but it does not prevent trauma from rocks, trees, cliffs or terrain traps, and it does not create a rapid exit after roads and slopes close.

Sources referenced