The fact that most offenders have accomplices at some point in their criminal career is curious, given the risks associated with criminal cooperation. McCarthy, Hagan, and Cohen () offered the first formal theory of the decision to co‐offend, which addressed explicitly the uncertainties attached to the decision to engage in group crime. They posited that when offenders experience adversity, they become more risk seeking and oriented toward the chance for potential gain, which essentially outweighs the uncertainties attached to criminal cooperation. McCarthy, Hagan, and Cohen's analysis of street youth offered some empirical support for their premise but left open many important questions. The current study uses data from two different samples of incarcerated felons in Nebraska (N = 321 offenders) and Colorado (N = approximately 1,120 observations nested within approximately 640 offenders) that provide information on different forms of economic adversity. Logistic regression models provide some evidence for the association between adversity and co‐offending, but they are inconsistent. In contrast, a preference for excitement is a consistent and powerful predictor of offending.