Based on 1990-1991 educational records from the South Carolina Department of Juvenile Justice, a retrospective study was designed. The sample consisted of 145 participants enrolled in the GED program. Participants classified as achievers composed the case group, and participants classified as nonachievers made up the control group. The independent variables of sex, race, age, educational level, IQ, and type of participant (mandatory or voluntary) were controlled. Adjusted odds ratios were used to show the relationship between the independent variables and voluntary participation. Multiple logistic regression was used as the model-building procedure to calculate the odds ratio, along with the 95 percent confidence interval for the association between voluntary participation and academic achievement in a prison literacy program. Results failed to show significant differences in academic achievement when program participants were there voluntarily or because of official mandates. The findings thus support mandatory literacy programs in the prison system. 3 tables and 11 references