OBJECTIVE:To investigate the effects of local labor market conditions and the availability of employer-sponsored health insurance on exits from the Medicaid program. DATA SOURCE:Data for this project come from a unique administrative database containing a 2 percent sample of all cases on California's Medicaid program in 1987 and a 2 percent sample of all new cases starting each year between 1987 and 1995. STUDY DESIGN:The results are estimated using a discrete duration model where the monthly exit probability is a function of demographic characteristics, local labor market variables, the probability of having employer-sponsored insurance, and fixed year and county effects. PRINCIPAL FINDINGS:Improvements in labor market opportunities (i.e., employment growth, wage growth, and increases in the availability of employer-sponsored health insurance) promote exits off the Medicaid program. A 2.5 percentage point increase in the availability of employer-sponsored insurance leads to a 6 percent increase in the probability that a completed spell lasts no more than 2 years. It would take a 2 percentage point decrease in unemployment rates or a 10 percent increase in average quarterly earnings to yield an equivalent increase in the likelihood of exiting Medicaid within 2 years. These effects are robust to the inclusion of county-level fixed effects and time effects. CONCLUSIONS:Medicaid expenditures and caseloads are sensitive to local economic fluctuations and secular trends in the availability of health insurance. Continued decreases in employer-based health insurance coverage will greatly increase the demand for public insurance coverage and the financial pressures on state governments.