In 2003 the Institute of Medicine called on health plans to collect data on their members' race and ethnicity as a foundation for improving the quality of care and reducing disparities. We describe the progress made toward collecting these data, the most commonly used data collection methods, and the challenges plans have encountered. We found that from 2003 through 2008, the proportion of plans that collected members' data on race and ethnicity doubled in the commercial market to 60 percent. It increased even more sharply to 94 percent and 83 percent, respectively, for plans covering Medicaid and Medicare Advantage enrollees. However, the scope of data collection varied greatly across plans, and data collection was an organizationwide initiative in a minority of plans. To fulfill the goals of recent legislation, including the Affordable Care Act, health plans will need to expand their efforts. Among other steps, plans and other key stakeholders should agree on uniform race and ethnicity categories, modify information systems to capture these data, and increase members' trust so that self-reported data-the most accurate data on race and ethnicity-can be gathered. R esearchers have documented racial and ethnic disparities in the use and quality of health care for more than two decades. In a landmark report, the Institute of Medicine recommended a comprehensive, multilevel approach to address disparities.1 A key recommendation was that health plans collect and report data on members' race and ethnicity as a necessary foundation for improving the quality of care and reducing disparities. 2 The collection of such data is uniquely important because, in contrast to health care providers, plans can obtain information about members who do not frequently use the health care system.