The hypothesis that job loss affects the incidence of clinically significant alcohol abuse is tested using panel data from the Epidemiologic Catchment Area project. Results suggest that the incidence of clinically significant alcohol abuse is greater among those who have been laid off than among those who have not. However, employed persons in communities in which total employment is unexpectedly low are at reduced risk of becoming alcohol abusers. The implications of the results for economic policy and for mental health services are discussed briefly.