The treatment group in the study came from participants in the Opportunity to Succeed Program as part of an impact evaluation. A total of 406 eligible offenders were randomly assigned to receive either OPTS case-managed services (the treatment group) or routine probation/parole supervision (the control group). The research cohort was recruited between mid- winter 1995 and September 1996. Baseline interviews were completed with 307 offenders by December 1996. CHAID analysis (Magidson, 1993) was used to examine between-individual variation, focusing on the relationships among hard drug use, full-time employment, and the percentage of street months that individuals participated in predatory or drug-dealing crimes. CHAID analysis is data driven. CHAID performs segmentation modeling, dividing a population into two or more distinct groups based on categories of the best predictor of a dependent variable. It then splits each of those groups into smaller subgroups based on other predictor variables. This splitting continues until no more statistically significant predictors can be found or until some other stopping rule is met. The analysis found strong relationships between drug use and full-time employment on predatory criminal behavior. A relationship was also found between full-time employment and drug-dealing behavior. Following Horney et al. (1995), the study used street calendar data in a hierarchical linear model (HLM) to examine both within-individual differences (e.g., changes in individual life circumstances in the year prior to incarceration) and between-individual differences (e.g., age of respondents). Results from the HLM analysis confirm Horney et al. findings that changes in life circumstances are strongly related to short-term criminal behaviors. Findings underscore the importance of the linkage between full-time employment and predatory and drug-dealing crimes. 1 figure and 2 tables