A randomized pretest-posttest-control group design study, replicated three times, each time with 24 subjects randomly selected and assigned to either a treatment or control group of 12 subjects each, was conducted to determine the effectiveness of a modified instructional strategy for accelerative and mastery learning of basic arithmetic by a male incarcerated population. A pretest was administered to test for initial group equivalence. The treatment phase lasted 6 weeks. Posttest for acquisition was administered, which was followed by a 2-week delayed test for retention. ANOVA allowed the rejection of the null hypotheses of no effect. The treatment resulted in higher posttest and retention test scores. (Author/MNS)