Stability & Test Slopes: One Tool to Understand the Snowpack

Welcome to The Safety Refresh, our monthly educational content series written by the Colorado Avalanche Information Center. 

The October 2026 Safety Refresh is made possible by support from Friends of CAIC partner, MTN FX.


Written by: Dylan Craaybeek, CAIC Central Mountains Backcountry Forecaster

Few snowpack tests are as widely used, or as widely misunderstood, as the test slope. A test slope is a small, inconsequential slope that is steep enough to avalanche but not large enough to cause any serious consequences if you get caught.

When used correctly, a test slope is a valuable tool that provides immediate, real-world feedback about avalanche conditions. When used incorrectly, it can reinforce dangerous assumptions, create a false sense of security, or even expose the tester to unintended avalanche risk. Test slopes are much better at telling us when conditions are unstable rather than suggesting conditions are stable.

The defining characteristic of a good test slope is that it is a small slope that will not produce a dangerous avalanche. There should be no terrain traps below the slope, and it should not be large enough to produce an avalanche that could bury you (D2 or larger) even if it releases full path (R4 or R5). You should also have a partner acting as a spotter from a nearby, safe location.

A common place to find useful test slopes is just above or just below a road. This was a small (about 10 feet tall) slope just below a road in the Elk Mountains, where I remotely triggered this avalanche while approaching the slope on December 27, 2021.

Test slopes are often an easier snowpack test and faster than digging a snowpit. On a properly selected, representative test slope, triggering an avalanche provides stronger evidence about instability than a snowpack test, which cannot fully replicate the complex processes involved in avalanche release.

Even larger snowpack tests such as the Rutschblock only examine a small section of the snowpack. A triggered avalanche, or even signs of instability such as shooting cracks or a collapse ("whumpf"), provides meaningful evidence of instability. Treat a slope that does not avalanche as only one additional data point rather than proof of stability.

An important limitation is that test slopes may not be the best tool if you are dealing with deeply buried weak layers or hard slabs where the slab is well supported on the flanks and bottom of the small slope.


Another good example of a test slope about twice the size of the CAIC forecaster. The lookers left side of the slope is not steep enough to slide but the lookers right side reached 38 degrees and was not large enough to produce a D2 avalanche. This avalanche happened naturally the day before this picture was taken. December 13, 2021 in the Elk Mountains.

An example of how I’ve used a test slope comes from a day of fieldwork in the Elk Mountains in March 2025. After monitoring snow conditions all winter, I knew that a recent storm buried a new weak layer and wanted to evaluate the sensitivity of the developing slab.

Throughout the day, we observed several shooting cracks and collapses on low-angle meadows, and long-column snowpack tests (ECTs and PSTs) produced unstable results. In part because of these observations, we avoided all large avalanche paths and instead found a small, 36-degree test slope on representative terrain (near treeline, north-facing slope) where an avalanche would have minimal consequences. 

My partner remained in a safe position while I traversed across the rollover, then skied onto the slope. It avalanched, knocking me off my skis and carrying me a few feet downhill. Although the avalanche was small and expected, it confirmed what the other observations had already suggested: steep, north-facing slopes at upper elevations with sufficient new snow to form a cohesive slab were the most dangerous slopes.

It also reinforced an important lesson: even a small avalanche can easily knock you over. If the terrain below the slope had included a cliff, rocks, or another terrain trap, even this small avalanche could have had serious consequences. That is why a good test slope needs a safe runout in case you are caught and carried.


Another example of a good test slope. This is a small, roughly 20-foot-tall slope in otherwise gentle rolling meadows. This avalanche occurred naturally and was sent to CAIC in a public observation near Mt. Sopris on February 7, 2026.

If a test slope avalanches, it is reasonable to assume that similar slopes could also avalanche and you should avoid them. If it doesn’t avalanche, you should not assume similar slopes are safe.

Confidence in the stability of a slope should come after considering many different pieces of evidence, including the avalanche forecast, weather trends, terrain characteristics, and any other relevant snow and avalanche observations encountered in recent days or weeks. 

Before exposing yourself to any test slope, consider the worst-case scenario and make sure that it is within your acceptable risk. Even a very small avalanche could still twist your knee or reignite an existing injury. Ultimately, safe travel in avalanche terrain depends on gathering and understanding a great deal of evidence from multiple sources. It should take overwhelming evidence to convince you that a slope is safe, but very little evidence to decide that it may be dangerous and should be avoided.


Snowmobilers remotely triggered this avalanche on the Grand Mesa during the 24/25 winter season. This is an example of what might not be a great test slope. The slope gently transitions from around 40 degrees down to about 10 degrees with no terrain traps at the bottom and looks relatively small when compared to those trees, indicating it could be a safe test slope. However, while this avalanche was only a D1 and not big enough to bury a person, had conditions been different, the avalanche could have broken higher up in the terrain and might have grown large enough to bury even a snowmobiler.