This talk summarizes the findings of a recent National Academy of Sciences study, sponsored by the Food and Drug Administration, concerning the treatment of missing data in clinical trials. The report focuses on approaches to the design and conduct of clinical trials to reduce the frequency of missing data, and principled statistical techniques for the analysis of clinical trials subject to missing data. For the latter, the scientific plausibility of assumptions underlying missing-data methods, and sensitivity analysis for deviations from missing at random, are particular points of emphasis.