Mountain slopes in the surrounding area of the Somma-Vesuvius volcano (Campania, southern Italy) are highly susceptible to shallow landslides due to the covering of ash-fall pyroclastic deposits and represent a relevant societal risk (e.g. deadly events of May 1998 in the Sarno area). Rapid shallow landslides are triggered by heavy rainfall, starting from little volumes of pyroclastic soil. For this reason, the knowledge of the distribution of the pyroclastic soil thickness is fundamental to analyze and understand landslide susceptibility. In literature, several soil thickness models are known for the catchment scale, but they usually account for hillslope processes which control the regolith formation and its distribution along slopes. Starting from a database of pyroclastic soil thickness data, distributed in a large mountain range around the Somma-Vesuvius, in this work three models were tested: two state-of-the-art models and a new model, which combine some features of the aforementioned two. The models were calibrated and validated with field data. The newly developed model achieved results that fit better the spatial distribution of ash-fall pyroclastic soils along the mountain slope. The outcome could be used as input parameter in landslide susceptibility assessments.