Every useful medicine has a therapeutic window, a dosing regimen in which it provides greater benefit than harm. Society faces a dilemma in its development of drugs for chronic conditions. We approve drugs use after studies with 1,000-10,000 persons. Then, a successful drug is used by 1,000,000 to 10,000,000 persons. A drug that harms 1 in 1,000 or 1 in 10,000 persons goes undiscovered in controlled clinical studies. As has been demonstrated for chronic use of non-steroidal anti-inflammatory drugs (NSAID), it can also go undiscovered through decades of general use. Dr. Zeger’s talk discusses one potential solution to this dilemma, randomized consumer trials (RConTs). The idea is to create an incentive for larger, simpler, randomized treatment trials after initial drug approval. The design and analytic issues with an RConT are discussed. The ideas are illustrated with designs for a vaccine and new NSAID trial.