This study explores whether the specificity of risk assessment for parasuicide repetition can be improved by measurement of two psychological variables, overgenerality of autobiographical memory & future fluency for positive events, in the immediate aftermath of the index parasuicide. In a longitudinal study, parasuicide patients deemed to be at high risk of repetition on the basis of sociodemographic factors were followed-up over a 12-month period. The most potent short-term predictor of parasuicide repetition was found to be scores on the Beck Hopelessness Scale, whereas in the longer term the number of previous parasuicides was the major predictor. It is concluded that for the heterogeneous parasuicide population as a whole, psychological variables are unlikely to improve upon the Beck Hopelessness Scale, sociodemographic risk factors & clinical interview in the prediction of parasuicide repetition. (26 refs.)