The handling of missing data is one of the most complex issues faced in the analysis of observational data. Missing data affect analytic precision and complicate the process of scientific inference by adding ambiguity and possibly bias to the results. Common solutions, including Complete Case analysis, Single Imputation, Multiple Imputation, and Likelihood-based approaches have pros and cons depending on situation, and all make assumptions that may be untestable. The purpose of this roundtable discussion is to promote discussion and sharing of ideas among practicing statisticians from government, industry, and academia on this challenging issue. The moderator will be William G. Hawkes, PhD, Statistical Scientist, Quintiles Real World and Late Phase Research division. Questions for Roundtable Participants: 1. How have you handled missing data in your own statistical work? 2. What are the strengths and weaknesses of the various approaches? 3. When faced with a missing data problem, what would you consider the best way to handle it?