The purpose of this study was to evaluate the effectiveness of rational-emotive therapy (RET) and rational-emotive imagery, RET alone, relaxation therapy with discussion of the roots of stress and a no contact control group in changing self reported affect and behavioral infractions in prisoners housed in a medium security correctional facility. This study extended the research on RET and rational-emotive imagery to a population that is characterized by its resistance to traditional treatment interventions. Unlike several studies in which only self-report measures were used to determine treatment outcome, this investigation incorporated a behavioral measure to determine if treatment had an effect on behavior, in terms of the number of infractions that prisoners committed. Forty male prisoners who ranged in age from 20 to 47 were chosen and randomly assigned to one of three treatment groups or to the no contact control group. Subjects were selected based on poor institutional adjustment as determined by behavioral or psychological problems. They were all felony convicts who were serving sentences from one year to life. Fifty percent were black, 28% hispanic and 22% were white. Their crimes ranged from burglary to murder. All subjects were given four adjustment and belief inventories, namely the Idea Inventory, the Personal Beliefs Test, the Eysenck Personality Questionnaire and the Profile of Moods Scale before and after treatment. The number of infractions reported before and after treatment was also calculated. The results, as determined by analyses of covariance and post hoc pairwise comparisons, indicated that prisoners in the RET plus rational-emotive imagery group and those in the relaxation therapy with discussion of the roots of stress reported improvement in "agreeability" and "clearheadedness" when compared with others in the no contact control group. Prisoners in the RET group showed a lower endorsement of self directed irrational "should" beliefs as compared with those in the control group. There were no significant reductions in the other irrational beliefs or in the number of infractions as a result of treatment.