The aim of this paper was to explore whether some individual characteristics (intelligence, impulsivity, adventurism, and empathy) predict membership in the group of criminal offenders with different intensities of prison sentence recidivism (non-recidivists, below-chronic recidivists, and chronic recidivists). The sample consisted of 1,689 prisoners (1,600 men) aged between 18 and 80 years, who came to the Diagnostic Centre in Zagreb in 2013 for a diagnostic treatment. For the purposes of this work, data on gender and educational level of prisoners was used, their age at the time of the diagnostic treatment, the data collected with the Eysenck impulsivity questionnaire (IVE), and data on intelligence collected with the Revised Beta test. In addition, data on earlier prison sentence was used to construct a measure of recidivism. Results of hierarchical binary logistic regression analysis showed that the impulsivity and empathy are significant predictors of the below-chronic recidivist group in an analysis with non-recidivists as a reference group, and that the impulsivity and adventurism are significant predictors of the chronic recidivist group in an analysis with non-recidivists as a reference group. Measured individual characteristics failed to distinguish between chronic recidivists and the below-chronic recidivist group. These results contribute to the understanding of repeated incarceration and provide important guidelines for its treatment and prevention.