The snow cover model SNOWPACK simulates snow stratigraphy for the locations of automatic snow and weather stations. Based on the stratigraphy, snow stability is predicted by calculating three stability indices. We verified the performance of the skier stability index SK₃₈ in order to develop a supporting tool for avalanche warning services. Since stability depends on snow stratigraphy, modelled grain type, grain size, hardness and density were first validated. The skier stability index SK₃₈ performed poorly in terms of identifying potential weak layers. By introducing a new stability formulation (SSI) that combined the SK₃₈ with differences of hardness and grain size across layer interfaces – known indicators of structural instability – the model performance was substantially improved. For manually observed flat field profiles, the SSI was significantly related with stability test results. At the regional scale, a statistically significant relation between predicted and verified stability was found. Stability patterns at the mountain range scale were reproduced. With these improvements the snow cover model SNOWPACK will develop into a valuable supporting tool for stability evaluation as done by avalanche warning services.