Results indicated that the recidivism predictions made on the basis of unstructured clinical judgments faired no better than chance at actually predicting the reconviction rates of offenders for any sexual, severe sexual, or violent offenses. On the other hand, both the Static-99 and the Static-2002 outperformed clinical judgment for predictions of sexual recidivism. However, the Static-2002 proved no better than clinical judgment for the prediction of severe sexual recidivism. Moreover, when tested for recidivism within 2 years, none of the three prediction models faired better than chance for any of the recidivism outcomes. The findings suggest that actuarial risk assessment measures are statistically more likely to predict sexual recidivism than unguided clinical judgments. Future research should focus on testing acute and stable risk factors for sexual offenders. Participants were 121 male sex offenders who were subjected to pretrial forensic assessments (FPE) in Denmark between 1978 and 1992. Psychiatrists employed at the two forensic psychiatric settings conducted the unguided clinical risk judgment. The Static-99 and Static-2002 were completed for each offender based on psychiatric reports and criminal records obtained from the Danish Central Crime Register. Data were analyzed using the Mann-Whitney test and Spearman’s rank correlation statistics. Tables, figure, references