The participants were all from the Kent County Correctional Facility Honor Camp in Grand Rapids. They had all been convicted of a felony or misdemeanor and received sentences of 1 year or less. They completed the Thomas-Kilmann Conflict Resolution Mode Instrument to assess five modes of conflict resolution: competing, collaborating, compromising, avoiding, and accommodating. The participants were then divided into three groups: experimental, educational, and recreational. All three groups met 4 times for 3 hours each during the same 30-day period. The experimental group received 12 hours of conflict-resolution training using the cognitive model. Another group took part in 12 hours of remedial education; the third group took part in 12 hours of recreation. All inmates were then retested using the same instrument. Results for the 55 participants for whom complete data were available were not conclusive, but they should be assessed with cautious optimism. The most promising outcome was the significant reduction in the competing mode of conflict resolution for the experimental group. In addition, the significant difference between the original norm group and the inmate participants in four of the five modes on both the pretest and post-test supports the literature finding that inmates need such training programs. Findings indicated that cognitive training programs have promise in the minimum-security correctional setting. Tables and 50 references (Author abstract modified)