Pair counseling is an intervention in which two previously unconnected children or adolescents of the same sex and age meet with a counselor to develop skills to improve social interactions. The structure of pair counseling allows for experiential learning and processing. By playing and interacting with a peer, the students learn to work through their conflicts in a natural setting under the guidance of a counselor, while working on problem-solving and using various negotiation strategies. The counselor facilitates interpersonal growth by working with students from various action orientations called "self-transforming" and "other-transforming." Some juveniles may be more demanding ("other-transforming"); whereas, others will be more shy and timid ("self-transforming"). In pair counseling, Selman has recommended pairing children of different action orientations so that by participating together in the activities and discussions, the students will move to a more acceptable balance in action orientation (Cohn and Selman, 1991). This study examined whether pair counseling can be adapted to a residential environment for juvenile offenders. Specific issues examined were whether moral dilemma discussion groups can be used effectively with dyads, whether pair counseling accelerates the release from training school, whether pair counseling can help incarcerated juvenile offenders form more appropriate peer relationships, and whether pair counseling has an impact on the recidivism rate of incarcerated juveniles. The sample for this study consisted of 28 male juveniles incarcerated at a North Carolina training school. The treatment group was composed of 14 students who were not randomly selected. The control group consisted of 14 male participants who were a mean age of 14 years 4 months. They were similar in age to the treatment group, and they were representative of the regular training school population. Pretesting of the participants occurred 1 to 2 weeks before the beginning of the 10-week intervention. Posttesting occurred the week after completion of the intervention. Evaluation findings support improvement in peer relationships. Overall, however, pair counseling did not stimulate moral reasoning for the majority of participants, and it did not have an impact on recidivism. 1 table, 49 references, and appended descriptions for various pair counseling exercises and activities