All subjects had completed at least two courses for credit and been students for a minimum of 8 months. Information on the subjects from correctional files encompassed educational information, criminal history, grade-point averages, and behavior on parole, as well as a number of other variables. The Statistical Information on Recidivism (SIR) score was also obtained, based on biographical and offense data calculated for each subject. The SIR score provided the prediction for post-release behavior. With this information, the study compared the predicted versus actual recidivism of the subjects. Based on their SIR scores, all the study subjects were judged to be at high risk for recidivism. The study assumed that if offenders of like backgrounds who were all assessed as high-risk individuals and who were all students in a particular prison education program showed significant differentiation in their conformity to the SIR prediction, it could reasonably be concluded that the cause was at least in part due to the impact of the education program. The study found that the men who took full advantage of the education program coalesced consistently in subgroups that surpassed their SIR prediction to a substantial degree. The study further isolated the factors that are particularly effective in education programs. These include the creation of a culture of academic achievement, a focus on students who show signs of improvement, the provision of opportunities for participation in program administration and governance, opportunities for extracurricular activities, and the encouragement and facilitation of continuing education after release. 4 tables and 13 references