Research over the past two decades has generally agreed that adolescent employment is associated with increased risk of involvement in criminal activity. Although this finding is quite uniform throughout the literature, the authors questioned whether methodological challenges are obscuring the true connection between adolescent employment and delinquency. Drawing on the responses of 4,168 youths who completed the 1997 National Longitudinal Survey of Youth (NLSY), the authors further examined the relationship between adolescent employment and delinquency. Bivariate analysis of the data confirmed the finding of increased risk of criminal involvement for adolescents who are employed. A second analysis was performed in which the relationship between the two variables was assessed after conditioning on a propensity score formed from basic demographic characteristics. These results revealed that the effect of employment on delinquency remained but was reduced by approximately 16.5 percent. A final analysis relied on the methods described by Rosenbaum and Rubin (1983), in which the sensitivity of the maximum likelihood estimate of the causal effect of work on crime to different sets of assumptions concerning potential confounding influences was tested. This analysis indicated that the effect of employment on criminal involvement is sensitive to variations in assumptions about the relationship of unobserved sources of criminal propensity and the decision to take a job. More research is needed on specifying why there is a positive association between adolescent employment and criminal involvement. The current findings suggest the need for care in the interpretation of that relationship. Tables, notes, appendix, references