Abstract The frequency of occurrence of hydrogeological disasters (HGDs), as well as the persistence of their impacts, are not evenly distributed. Hazardous areas, by definition, are more prone to extreme events, while in densely urbanized regions, the impacts of these events tend to be more severe. The objective of this study is to investigate statistical relationships between urban and natural environment features and HGD occurrences. Taking Italian provinces as a comprehensive case study, we assessed the coefficient of determination, the χ 2 test, and the p-value to determine the degree of statistical correlation between impact indicators and 57 hazard/risk/land management indicators, such as extension of at-risk areas or soil sealing. We discovered that HGDs persistence and frequency correlate best with an indicator describing the amount of soil sealing (i.e. urbanized soil) in medium-hazard areas. Building on that, a further dynamic analysis was carried out to investigate whether soil sealing trends changed significantly after the provinces were struck by HGDs. Our findings hold significant implications, challenging current policy norms. European directives and Italian national laws impose strict development restrictions in ‘high-hazard’ areas, but generally allow for urbanization in ‘medium-hazard’ areas, with only minor limitations. Moreover, a paradoxical positive urbanization trend is observed in the most sensitive areas, greater than in safer areas and generally unchanged after HGDs. This outcome highlights a critical gap in risk perception that reflects into territorial planning, decision-making processes, and existing policies.