To evaluate the relative effectiveness of traditional versus computer managed instruction (CMI) basic skills programs for incarcerated adults, 359 male inmates from three traditional self-paced and three PLATO/CMI programs were given pretests in the Test of Adult Basic Education (TABE) and the Tennessee Self-Concept Scale (TSCS). Following three months of basic skills instruction, 47% of the original sample were administered TABE and TSCS posttests. Results showed significant gains for all students in language, math, and reading, with the greatest gains in language and the least in reading comprehension. Problems in administering and scoring the self-concept scale invalidated that test's results. Findings failed to indicate a way of using information on student age, ethnicity, type of prison sentence, and institution's security level to predict which students would gain the most from either classroom or CMI instruction. (MM)