AIMS: The association between drunk driving (DD) and/or risky driving (RD) offences and subsequent hospitalization and mortality was studied during a 25-year period. METHODS: Information about drinking habits and psychosocial factors for the 8122 conscripts from Stockholm County in 1969-1970 was linked to register data on hospitalization, mortality, DD and RD. RESULTS: Analyses comparing background characteristics of DD and RD groups showed that the prevalence of problem behaviour and drug use was highest in the RD group. The relative risk (RR) for hospitalization after DD and/or RD was significantly elevated in multivariate logistic regression analysis for all the studied diagnostic categories (alcohol diagnoses, narcotic diagnoses, suicide attempts, psychoses, E-codes, all diagnoses), and was especially high for alcohol (RR = 7.2) and diagnosis of drug misuse (RR = 9.2). The RR of all hospitalization was 1.5 for the DD group, 1.8 for the RD group, and 1.9 for those who had been sentenced both for drunk driving and risky driving (DRD), all of which were significantly increased. The RR of death was significantly elevated in all three groups. CONCLUSIONS: The results show a significantly increased risk of hospitalization and mortality both in the DD and the RD group. From a public health and traffic safety perspective, this implies a need for developing and implementing better prevention strategies.