Using longitudinal data on the criminal careers of a group of high-risk men and women (N 540) who were institutionalized in a Dutch juvenile justice institution in the nineties, this article addresses the effects of (un)employment on crime. Results show that, for both men and women, employment rates are below average and stability in employment is low. Nevertheless, random effects models consistently show employment to reduce the estimated number of convictions for both men and women. Employment duration has an additional effect on crime, but only for men. Unemployment duration increases the estimated number of convictions for women, while slightly decreasing them for men. © 2012 The Author.