Administrative data on a 10-percent random sample of incarcerated male prisoners in November, 1975 provided the basis for this path analysis. High custody grade was found to have a strong direct path with the granting of work release, and marital status had a weaker path to release. Court recommendations had an indirect path to work release by positively affecting custody grade. Older prisoners and those with less serious crimes were more likely to receive court recommendations for the granting of work release. For short-term prisoners, strong direct paths between court recommendations and both work release status and custody grade were observed. Work release appeared typically to be a sanction for successful performance in a long career in a prison system and to be affected directly by court recommendations for some prisoners at time of sentencing. Work release may also serve as a device which forestalls imprisonment for a small group of prisoners and may serve as a mechanism for reintegrating prisoners into society prior to their release. Tabular data, diagrams, and 51 references are provided. (Author abstract modified)