Suicide on railway networks comprises a serious public health problem. However, the geographical distribution and the environmental risk factors remain unclear. This study analyzed the geographic distribution of railway suicides in Belgium from 2008–2013 at the level of a railway section (average length of 3.5 km). Principal component analysis (PCA) identified three groups of correlations that helped explain the variance of railway suicide. The three groups are related to characteristics of urban spaces, psychiatric facilities, and railway traffic density. Based on the PCA results, the study found four types of railway sections. The density of railway suicide was average and low in the urban and rural/industrial sections, respectively. However, it was high in the suburban sections and the sections close to psychiatric facilities. As the geographical proximity of a psychiatric facility comprises a specific risk factor for suicide on railways, preventative measures should target these sections and establish collaborations with psychiatric facilities. The typology of locations found in this study constitutes crucial information for national and local suicide prevention on the Belgian railway network. ; SCOPUS: ar.j ; info:eu-repo/semantics/published