CONTINUED INCREASES IN THE NUMBERS OF UNINsured US citizens, increasing premiums for those who are insured, and the debate over prescription drug coverage all have served to thrust the issue of access to health care on the national radar screen once again. Since the goal of universal health insurance seems less and less likely to be realized in the near future, much of the discussion has focused on the so-called safety net, which is the loosely knit health system that cares for those outside the traditional health care marketplace, who are primarily poor and uninsured. The mere existence of the safety net, let alone our nation’s dependence on it, is evidence that the market does not distribute health care in a socially equitable manner. This often ad hoc (and unheralded) system is expected to provide access to care in geographically underserved areas and to address other barriers to care, such as social and cultural factors. With the increasingly diverse US population, additional factors include whether the patient and clinician can communicate in the same language and whether they understand and are comfortable with each other’s cultures. Because of the intrinsic links among access to care, quality, and outcomes, the safety net ideally should go beyond simply providing access to provide high-quality care to those it serves. The safety net functions against a backdrop of uninsurance and underinsurance and the inability of many who need care to pay for it. These problems are not distributed evenly across the population. Rural residents are more likely than those living in metropolitan areas to be uninsured. Furthermore, recent increases in rates of uninsurance have affected population groups differentially. Between 1994 and 1998, the number of uninsured whites increased by 3.4% while the number of uninsured blacks and Hispanics increased by 17.4% and 21.5%, respectively. In addition to its shortcomings in providing affordable access to care, the market has been unsuccessful in distributing health care professionals to geographic areas according to need. Federal efforts, through community health centers and the National Health Service Corps (NHSC), have helped mitigate these market failures. In this issue of THE JOURNAL, Pathman and colleagues present national data about the extent of nonfederal efforts to improve the distribution of the clinician workforce. They found that in 1996, 41 states had scholarship or loan repayment programs or other mechanisms to encourage medical practice in geographically underserved areas. They report that the combined number of clinicians who practice in underserved areas as a result of such programs is at least equal to that of the federal government’s NHSC. While the growth of such programs is undoubtedly a positive development, it must be put into context. The NHSC addresses only a small fraction of the unmet need, and even doubling that effort by adding those practitioners recruited by state programs still results in an inadequate number of practitioners needed for underserved areas. Indeed, the NHSC estimates the unmet need for clinicians at roughly 18000, even after accounting for the practitioners identified in the study by Pathman et al. Even if the safety net had an optimal geographic distribution of health care providers, concerns about its longterm viability would remain. The Institute of Medicine recently identified 3 major factors that will hinder the ability of clinicians within the safety net to deliver high-quality care. First, these clinicians are likely to be disproportionately affected by the growing numbers of uninsured persons. Second, traditional subsidies that helped finance the safety net are eroding. Third, the rapid shift to Medicaid managed care has had adverse consequences for many providers within the safety net. While there is no direct evidence that the clinicians in programs studied by Pathman et al do, in fact, care for uninsured patients, it is well recognized that the areas in which they practice are more likely to have higher rates of uninsurance, so the demand to provide uncompensated or undercompensated care is likely to be greater than it is for other practitioners. If, in addition, the institutions that support these clinicians are increasingly under fiscal stress, their ability to develop or sustain an infrastructure to provide high-quality care is also threatened. Beyond geographic distribution, social and cultural factors also have long been recognized as critical components of access to care. Sex, language, and race/ethnicity are all