The FDA initially requires add-on studies with new AEDs due to risk of seizures during the placebo phase. The author argues that bipolar research warrants similar add-on studies to prove efficacy because the risk of suicide & increased long-term morbidity in the bipolar population is as significant as the risk of seizures in the epilepsy population. Although the number of patients needed to prove statistical efficacy would increase, the safety of such research would also markedly increase. The author also concludes that with the risk of suicide during bipolar research, ethical considerations require increased frequency of patient contact with a significant other co-signing the informed consent for research & serving as a contact for the coordinator. (59 refs)