The content areas of the Criminal Attribution Inventory (CRAI) include psychopathology, personal, alcohol, victim, society, and randomness. One of the purposes of the CRAI is to evaluate pretreatment and posttreatment differences in the perception of criminal blame. Given the strategies of the CRAI to reduce response styles and external threats to validity, the current study examined the relationship between the CRAI and the Impression Management scale of the Balanced Inventory of Desirable Responding (BIDR) developed by Paulhus (1984). The 5 samples consisted of 143 incarcerated male offenders, 26 sexual offenders who were participating in a sexual offender program, 70 first-year undergraduate students who were enrolled in an introductory psychology course, 61 male offenders who were participating in a community program that targeted antisocial attitudes, and 65 male offenders who were applying for a highly desirable job within a minimum-security institution. The contention that the CRAI will have a minimal relationship with socially desirable responding was supported. Other studies with incarcerated offenders have typically shown stronger correlations between socially desirable responding and self-report measures. In a "fake-good" testing situation, the faking index had appropriate sensitivity, specificity, and positive and negative predictive power in identifying "fake-good" responses. The CRAI's instructional set of general observation compared to self-representation apparently limits socially desirable responding and also allows the CRAI to be administered to nonoffender populations. 2 tables and 40 references