Background: First episode psychosis (FEP) patients have an increased risk for violence and criminal activity prior to initial treatment. However, little is known about the prevalence of criminality and acts of violence many years after implementation of treatment for a first episode psychosis. Aim: To assess the prevalence of criminal and violent behaviors during a 10-year follow-up period after the debut of a first psychosis episode, and to identify early predictors and concomitant risk factors of violent behavior. Method: A prospective design was used with comprehensive assessments of criminal behavior, drug abuse, clinical, social and treatment variables at baseline, five, and 10-year follow-up. Additionally, threatening and violent behavior was assessed at 10-year follow-up. A clinical epidemiological sample of first-episode psychosis patients (n = 178) was studied. Results: During the 10-year follow-up period, 20% of subjects had been apprehended or incarcerated. At 10-year follow-up, 15% of subjects had exposed others to threats or violence during the year before assessment. Illegal drug use at baseline and five-year follow-up, and a longer duration of psychotic symptoms were found to be predictive of violent behavior during the year preceding the 10-year follow-up. Conclusion: After treatment initiation, the overall prevalence of violence in psychotic patients drops gradually to rates close to those of the general population. However, persistent illicit drug abuse is a serious risk factor for violent behavior, even long after the start of treatment. Achieving remission early and reducing substance abuse may contribute to a lower long-term risk for violent behavior in FEP patients.