OBJECTIVE: To examine gender differences in prediction of long-term outcome in first episode psychosis (FEP). METHOD: Eighty-one male and 72 female FEP patients were compared regarding the sensitivity and specificity of the Predictive Rating Scale (PRS). The contributions of pre-admission clinical and socio-demographic characteristics to a poor 5-year outcome were analysed for males and females separately. Gender differences in the relations between predictors and outcome were examined using the equality of correlation comparing correlation coefficients. RESULTS: The sensitivity of the PRS was significantly better for males than for females. The following items: 'the highest Global Assessment of Functioning (GAF) the year before first admission