were positively or negatively correlated; and b) to investigate potential predictors/mediators of suicide risk alone, violence risk alone, and risk for both suicide and violence. Suicide risk and violence risk were considered outcome variables. The predictor/mediator variables were eight ego defense mechanisms and 13 dysfunctional personality styles (both have been linked to aggressive risk). Seventy-four male alcoholics were administered questionnaire measures of suicide risk and violence risk, defenses, and disordered personality functioning. Results indicated that suicide and violence risk were significantly positively correlated. Three sets of predictor/mediator variables were identified: those predicting violence risk alone, those predicting suicide risk alone, and those predicting combined aggressive risk (risk for both suicide and violence). Connections obtained between predictor/mediator variables and outcome variables were explained in the context of existent clinical theory.