The treatment consisted of eight weekly small-group discussions of sociomoral dilemmas. The group participants initially differed in both sociomoral stage and dilemma opinion. Of the dilemma-discussions subjects, 87.5% were pretested at Modal Stage 2 (in terms of Kohlberg's moral judgment stages) and shifted to Modal Stage 3 on the posttest, whereas only 14.3% of the Modal Stage 2 control subjects did so. Pretest Modal Stage 3 subjects remained at Stage 3 on the posttest. Across the experimental groups, no difference was found between groups where a consensus was required for the discussion and groups having no consensus requirement. Investigations of the possible in-program and postrelease behavioral ramifications of the dilemma-discussions treatment are recommended. (Publisher abstract)