On the basis of Murray's theory, it was hypothesized that the interaction between internal needs and environmental influences (press) would be the best predictor of rehabilitation. Dependent measures of rehabilitation applied to the postincarceration period were number of months before rearrest, number of months before applying to social services for financial support, and number of months before leaving the narcotic antagonist program. The independent variables selected were press for addiction, defined as one or both parents having been addicted to alcohol or other drugs through the subject's 16th year; press for family support, which was defined as the absence of one or both parents through the age of 16; deference; succorance; and autonomy. High or low levels of need were operationally defined in terms of whether subjects' scores on the relevant items of the Edwards Personal Preference Schedule were above or below the median score for the entire sample. Subjects were 68 male inmates in the Nassua County Correctional Center with histories of drug addiction who had participated in a narcotic antagonist work-release program. Data were evaluated with stepwise regression analysis and a nonparametric technique. The study concluded that the usefulness of considering the interactions between predictors of rehabilitation in incarcerated heroin addicts remains to be documented; however, the failure to document the hypotheses was attributed to the overall difficulty in predicting rehabilitation rather than to the inappropriateness of the interactional model. Analysis techniques are appended, and tabular and graphic data are provided, along with about 60 references. (Author abstract modified)