The availability of large state and federally run administrative health-care databases provides potentially comprehensive population-wide information that can dramatically impact both medical and health-policy decision-making. Specific opportunities and important limitations exist with all administrative databases based on what information is collected and how reliably specific data elements are reported. Access to patient identifiable-level information can be critical for certain long-term outcome studies but can be difficult (although not impossible) due to patient privacy protections, while more easily available de-identified information can provide important insights that may be more than sufficient for some short-term operative or in-hospital outcome questions. The first section of this paper by Sarah K. Meier and Benjamin D. Pollock discusses Medicare and the different data files available to health-care researchers. They describe what is and is not generally available from even the most granular Medicare Standard Analytic Files, and provide an analysis of the strengths and weaknesses of Medicare administrative data as well as the resulting best and inappropriate uses of these data. In the second section, the Nationwide Inpatient Sample and complementary State Inpatient Database programs are reviewed by Steven M. Kurtz and Edmund Lau, with insights into the origins of these programs, the data elements that are recorded relating to the operative procedure and hospital stay, and examples of the types of studies that optimally utilize these data sources. They also detail the limitations of these databases and identify studies that they are not well-suited for, especially those involving linkage or longitudinal studies over time. Both sections provide useful guidance on the best uses and pitfalls related to these important large representative national administrative data sources.