In this paper, we use data from the Health Survey of England to show that problem drinking is negatively and significantly associated with the probability of being in work, once the endogenous relationship between these outcomes is accounted for. Being a problem drinker leads to a substantial reduction in the probability of working by between 7 and 31%, the former figure being roughly equivalent to the positive effect of having a degree relative to no qualifications in our data. This finding is robust to a variety of identifying restrictions and definitions of problem drinking. Moreover, we find that problem drinking defined by the observed psychological and physical symptoms of alcohol is an important predictor of employment, and allows for the fact that individuals differ in their tolerance or susceptibility to alcohol. Our results suggest that there may be important labour market benefits from public health policies aimed at the prevention and treatment of problem drinking. Copyright © 2003 John Wiley & Sons, Ltd.