Study data came from 2,495 persons released from Federal prisons during 1972 and 432 high-risk male offenders released from Maryland's State prisons. The first sample consisted of an older, predominantly white, hardened criminal population, while the Maryland offenders were disproportionately black and young and had poor previous employment experiences. In the Federal sample, more stable preprison employment was associated with lower postprison recidivism, but the relationship was weak. Other background variables and measures of the certainty and severity of punishment explained the variance in recividism just as well. However, increased certainty of punishment increased recividism rather than lowered it. In the Maryland sample, preprison employment experiences had weak effects, but postprison employment significantly affected postprison rearrest. Higher wages and more hours worked were strongly associated with lower recidivism. Taken together, the samples indicate that what happens in the labor market after release from prison is what matters in determining a return to crime. Data tables and a list of 29 references are included.