Beginning in 1990, Medicare set volume standards, and subsequent payment updates for physician fees depend on the difference between actual volume and the standard. Establishing a VPS (volume performance standard) policy requires choices along three dimensions: the risk pool, the scope and nature of the standard, and the application of the standard. This study reviews the literature and the experience of other countries to analyze the strengths and weaknesses of these alternative choices. The authors conclude that VPSs will be most effective in controlling expenditures and changing physician behavior if they are defined using subnational geographic units, all Medicare services (Part A and Part B), and per-capita utilization targets. Other countries have successfully controlled costs with expenditure ceilings, though there is limited evidence as to how practice patterns have responded to these ceilings. However, capitated payment systems have been demonstrated to be effective in lowering health care use, and the Health Care Financing Administration should continue to encourage prepaid plans.