Overall, the findings indicate that the CSS in an internally consistent tool for the assessment of criminal sentiments among sex offenders; however, it predicts nonsexual reoffending but not sexual reoffending. The CSS was a robust predictor of nonsexual violent and nonviolent reoffending; and it correlated significantly with the General Statistical Information on Recidivism Scale, which assesses nonsexual reoffending risk. Similar analyses, however, failed to show that the CSS could not reliably predict sexual reoffending, nor did it correlate with a tool that assessed risk for sexual reoffending, i.e., the Static 99. Using the CSS at pretreatment and posttreatment in an accredited sex offender treatment program produced changes in the CSS total and subscale scores in the expected direction; however, with a control group, researchers were unable to attribute the observed changes to the effects of treatment. Suggestions are offered for future research. The 72 sex offenders who were treated in a sex offender program completed the CSS before and after treatment, and they were followed up for approximately 3 years. 3 tables and 32 references