Professional risk assessment provides estimates of future risk in terms of nature, frequency and severity as well as may determine treatment services required of an offender under parole or probation supervision. This study examined the predictive validity of the HCR-20 (Webster, Douglas, Eaves, & Hart, 1997) in a Lithuanian forensic context.

This research includes a sample of criminal offenders and a sample of forensic psychiatric patients (a total of 118 participants). The HCR-20 was coded on the basis of file information.

The mean HCR-20 scores were: Total score = 14.96 (SD = 6.56), H-scale = 8.42 (SD = 3.61), C-scale = 2.57 (SD = 2.16), R-scale = 3.89 (SD = 2.55). ROC analyses of the HCR-20 subscales showed AUC's of .72 for the H Scale, .69 for the C Scale and .58 for the R Scale.

The HCR-20 total score and final risk judgments were significantly more accurate in predicting violent recidivism (p < .05). Results indicate the predictive ability of the instrument may be maximized when judgments of final risk are used rather than an actuarial approach wherein individual risk factors are summed.