Studies of inmate deviance have focused on two groups of variables: pre-institutional variables such as demographic and socioeconomic characteristics and individual and aggregate institutional variables. The main problems with these studies are improperly specified models, inadequate data collection techniques, and analyses of inappropriate dependent variables. Nevertheless, most of the studies suggest that inmate deviance may be reduced by means of closer supervision of inmates who are younger, who committed violent crimes, who are emotionally or mentally unstable, who come from urban backgrounds, who have not been incarcerated before, who have shorter sentences, and who are closer to the middle stages of their sentences. In addition, increasing the frequency of visits between inmates and their family members may help reduce the likelihood of deviant behavior. Finally deviance may be less prevalent in less crowded facilities. Further research should examine th relative importance of structural and individual variables in relation to problem behavior. 75 references (Author abstract modified)