Sean Tunis and colleagues provide an excellent critique of current federal activities to assess new medical technology. These efforts generally do not involve primary data collection but rather reflect attempts to better synthesize existing information, to make conditional coverage decisions based on the data, and to increase coordination among government agencies. Many challenges remain on analytical, logistical, legal, and political fronts. Researchers and analysts should be more precise about what "rapid learning" means and strive to measure performance. Efforts are also needed to prioritize research, to communicate it to decisionmakers, to involve stakeholders in the process, and to include cost-effectiveness information.