Data collected on 1,025 individuals included personal, criminal, and employment background; short (6 month) and long-term (2 year) recidivism; and employment services participation. Subjects were divided into three groups: experimental subjects who received more individualized and intensive followup services, a control group that received normal program services, and a nonrandom comparison group that received no services. Results of least-squares regression analysis failed to show significant differences in recidivism among subjects receiving intensive versus normal services, but program participants at the Chicago site did show significantly lower recidivism rates than those receiving no placement. Three of five demographic variables were significant predictors of recidivism: age (inversely related to recidivism in Boston and San Diego), race, and marital status. In general, results suggest that ex-offender employment programs have only a small impact on recidivism rates. Tabular data, an appendix, 75 footnotes, and 89 references are provided. (Author abstract modified)