This paper examines forecasting activities among Medicaid agencies in the fifty United States, Washington, D.C., and five U.S. territories (American Samoa, Guam, Puerto Rico, Northern Mariana Islands, and Virgin Islands). Most frequently, studies of state or local forecasting practice focus on revenue forecasting. There are several reasons why comparison of state Medicaid forecast practice may be better than comparison of state revenue forecasting practices. First, there is no consistent reporting of state revenue estimates. States make forecasts when it suits them and report them in a manner that is satisfactory to their governors or legislatures. Collection of data through national organizations such as the National Association of State Budget Officers is not so rigorous as to assure that reported data are comparable. In contrast, Medicaid agencies must report their expenditure estimates to the federal government using the federally specified HCFA-37 form once a quarter beginning roughly 30 months before the end of each federal fiscal year. Second, determining the accuracy of state revenue forecasts relies on the validity of state reported differences between planned and actual expenditure. States may be politically motivated to report these data in a favorable manner. By contrast, Medicaid forecasts, can be compared with accounting data as reported on a federal report known as the HCFA-64. While these data may not be bias free,3 biases are likely to be small and similar from state to state.