Subjects were all male inmates released from New York State correctional facilities during the period from July 30, 1982, through September 1, 1983. Included in the group were all inmates who experienced a psychiatric commitment during the instant incarceration (n=150). A stratified sampling design selected inmates who had not required psychiatric commitment during imprisonment from the same exit cohort (n=400). A weighting equation was then applied to give each element in the sample its corresponding weight in the population, thereby making the comparison group representative of the general prison population. The legal variables were number of prior imprisonments, prior arrest for violent crime, use of violence in current offense, conviction for drug-related offense, history of drug use, and annual prison infraction rate; social and community support variables were marital status, education, area of conviction, family support, and occupation; and extralegal variables were race and age at release. These variables, along with psychiatric hospitalization history, were examined for their impact on the parole decisionmaking process and the percentage of maximum sentence served. The results from logistic regression and multiple regression show that psychiatric hospitalization history significantly reduced the likelihood of receiving parole and increased the percentage of maximum sentence served. This relationship persisted once other variables had been controlled. Possible explanations and policy implications are discussed. 3 tables and 50 references