About 1 in every 20 employees experience depression in the U.S. A recent study estimated that in 1990 the economic costs of depression amounted to $43 billion, with absenteeism alone contributing $12 billion. OBJECTIVES: The study used 1996-1997 data collected in crime-ridden and low-income neighborhoods of Miami, Florida to (1) examine the relationship between depression and employment, (2) conditional on being employed, estimate the effect of depression on annual weeks worked, (3) examine the robustness of the model estimates to the co-morbid effects of drug use. METHODS: The labor supply measures included employment in past 30 days and number of weeks worked in past 12 months. The analysis estimated a univariate probit model of employment as well as a bivariate probit model of depression and employment, which accounted for the possible correlation between the unobserved determinants of depression and employment. The annual weeks worked specification was estimated by a standard Tobit and an instrumental variable Tobit model, which, besides the censoring of the observations, accounted for the possible endogeneity of depression. RESULTS: Results indicate that depressed individuals had a 0.18 lower probability of employment relative to non-depressed individuals in the univiariate probit model and a 0.15 lower probability in the bivariate model. Both standard and IV Tobit models found that depression significantly reduced the number of annual weeks worked by 8 weeks. Co-morbid drug use significantly contributed to the estimated effects of depression. CONCLUSIONS: Prevention and/or treatment of mental health problems such as depression may yield economic benefits by promoting employment and enhancing labor supply. While expansion of mental health services may be particularly beneficial to the unemployed, employers may find it economically efficient to allocate more resources to stress management training and job redesign.