Public attitudes toward sex offenders have contributed to the passage of a number of sex offender-specific laws. Following these laws, researchers developed the Community Attitudes Toward Sex Offenders (CATSO) scale, yet there is debate over the most appropriate factor structure, and no studies to date examine its predictive utility. The current study tested the CATSO factor structure, and explored construct and predictive validity in a community sample. Results included identification of (a) a 2-factor model consisting of capacity to change and social isolation factors; (b) expected construct validity patterns between capacity to change beliefs with political ideology, attribution of blame toward offenders, and sentencing decisions; (c) utility of capacity to change beliefs, but not social isolation beliefs, in predicting sentencing recommendations, placement decisions, and blame attribution. Implications are discussed concerning potential bias in sex offender-related trials and policy, as well as recommendations for application of the CATSO in future research.