There is a lack of insight into the relationships between violent ideation, violent behavior, and early, particularly attenuated, psychosis. Our aims were to examine the relationships between baseline violent behavior and violent ideation and outcome violent behavior and conversion to psychosis in at-risk individuals. We longitudinally assessed 200 individuals at clinical high risk for psychosis for violent ideation and violent behavior using the Structured Interview for Psychosis-Risk Syndromes (SIPS), and rated these according to MacArthur Community Violence categories. Fifty-six individuals (28%) reported violent ideation at baseline, 12 (6%) reported violent behavior within 6 months pre-baseline, and 8 (4%) committed acts of violence during the follow-up time period. Information about violent ideation was obtained only by indirect, but not direct, inquiry about violent ideation. Both violent ideation and violent behavior at baseline significantly predicted violent behavior (RR=13.9, p=0.001; RR=8.3, p=0.003, respectively) during follow-up, as well as a diagnosis of psychosis (RR=2.3 and 2.4, respectively; both p<0.001), independent of more than 40 clinical and demographic variables. The targets of the subjects' violent ideation at baseline were completely different than their subsequent targets of violent behavior. Violent behavior occurred within 7 days (SD 35 days) of a diagnosis of syndromal psychosis. These data suggest that checking carefully for violent ideation and behavior in clinical high-risk patients is essential, as these have predictive value for conversion to psychosis and likelihood of violence in the future.