As the first systematic evaluation of an American work-release program, this investigation employed data on all male prisoners placed on the program in North Carolina between 1957 and 1963. Outcome was measured against the variables affecting selection for work-release. The study found that factors affecting parole performance after work-release include: (1) length of term served before work-release, (2) length of time on work-release, (3) skill level of work-release job, (4) continuation of work-release job into parole, and (5) relationship between work-release and parole locale. A sample of orthodox parolees was compared with the paroled work-releasees to disclose possible differences in parole performance and problems raised during supervision. The study concluded that the work-release experience generally brought fewer problems for the parole supervisor, in terms of transgressions, mobility, and family problems. The problems raised by the work-release parolee were more likely to be associated with behavior consistent with the ultimate purpose of parole. [Not available in hard copy due to marginal legibility of original document.] (BH)