Substance use disorders are overrepresented in incarcerated male populations. Cost-effective screening for alcohol and substance use problems among incarcerated populations is a necessary first step forward intervention. The Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST) holds promise because it has strong psychometric properties, requires minimal training, is easy to score, is available in the public domain but, because of complicated skip patterns, cannot be self-administered. This study tests the feasibility, reliability, and validity of using computer-administered self-interviewing (CASI) versus interviewer-administered interviewing (IAI) to screen for substance use problems among incarcerated men using the ASSIST. A 2 × 2 factorial design was used to randomly assign 396 incarcerated men to screening modality. Findings indicate that computer screening was feasible. Compared to IAI, CASI produced equally reliable screening information on substance use and symptom severity, with test–retest intraclass correlations for ASSIST total and substance-specific scores ranging from 0.7 to 0.9, and ASSIST substance-specific scores and a substance abuse disorder diagnosis based on the Structured Clinical Interview (SCID) were significantly correlated for IAI and CASI. These findings indicate that data on substance use and symptom severity using the ASSIST can be reliably and validly obtained from CASI technology, increasing the efficiency by which incarcerated populations can be screened for substance use problems and, those at risk, identified for treatment.