This paper presents a choice experiment, which values reductions in mortality risk on Alpine roads. These roads are on one hand threatened by common road hazards, on the other hand they are also endangered by natural hazards such as avalanches and rockfalls. Drawing on choice data from frequently exposed and barely exposed respondents, we are not only able to estimate the VSL but to explore how the respondents differ in their individual willingness-to-pay depending on personal characteristics. To address heterogeneity in preferences for risk reduction, we use a non-linear conditional logit model with interaction effects. The best estimate of the VSL in the context of fatal accidents on Alpine roads is in the range of €4.9–5.4 million with distinct differences between the urban and the mountain sample groups. We find the VSL to be significantly altered by socio-economic factors but only marginally altered by the type of hazard.