The sample was drawn from the Baltimore Living Insurance for Ex-Prisoners (LIFE) experiment and consisted of males released from Maryland State prisons to the Baltimore metropolitan area. All had low financial resources, were repeat property offenders, had no known history of alcohol or narcotics abuse, and had not been on work release for more than 3 months. Two variants of the economic model of crime were estimated using maximum likelihood techniques. Independent variables included measures of the certainty and severity of punishment as well as proxies for legitimate opportunities. In particular, measures of expected wages and prior employment experience were considered. In one specification, the dependent variable was the probability that an offender was ever rearrested in the first year following release from prison; increased wages were found to be associated with significantly lower rearrest probabilities. A second specification considered monthly 'survival' probabilities; the dependent variable was the probability that the individual was not rearrested. In every month but the first, the average weekly wage was positively related to postprison success and was significant at the 1-percent level. In the first month the level of significance dropped to 10 percent, but was still positive. Virtually no evidence was found for the contention that increased certainty or severity of punishment increases the postprison survival rate; however, strong and consistent support emerged for the contention that better wages reduce crime. Included are 14 references. (Author abstract modified)