Appropriate use of existing data, and creative uses of existing data collection mechanisms, have been and will continue to be crucial to assessing safety of marketed drugs. The goal of such assessments will be improvement of healthcare decision-making, which also requires a full understanding of effectiveness in actual clinical settings, so that benefits, as well as risks, can be assessed. With this goal in mind, the strengths and limitations of currently available administrative data to address questions of comparative safety (and effectiveness) will be explored, and additional ways in which the existing infrastructure for these databases might be used to support further data collection efforts, perhaps more specifically targeted, will be proposed.