The data were obtained from self-administered surveys completed by 1,531 college students in the fall of 1996. The students were enrolled in nine postsecondary institutions in eight States. The sample was drawn from students in introductory social science general education courses. Students who were employed comprised 62 percent (n=907) of the sample. A 95-item instrument was used to assess individual demographics, social and leisure experiences, community structural and contextual variables, and elements of the respondent's employment status and job attributes. Also included were a wide range of self-reported illegal activities and substance uses. The dependent variable was whether or not respondents had stolen from their employers. Independent variables included job characteristics, alcohol and drug use, illegal actions, and demographic characteristics. The job characteristic category was intended to measure such routine-activity concepts as opportunities, interactions with coworkers and customers, as well as when and where work actually occurred. The students who stole from their employers were found to be more likely to be heavy public drinkers, hold multiple jobs, handle cash, and have previously engaged in thieving behavior. Also, their past criminal behavior apparently had been relatively serious, since these students were more likely to have spent time in prison. Routine-activity theory suggests that criminal events are more likely to occur when there are offenders who are inclined to make use of available opportunities. Thus, this study shows that when individuals who have shown a prior criminal inclination are confronted with job-related situations that offer opportunities for illegal activity, they will act on such opportunities. Alcohol use is also related to employee theft in this study. Suggestions for future research are offered. 3 tables and 71 references