Practitioners of clinical trials have a responsibility to ensure that patients' participation in research be informed and voluntary. This responsibility implies that we should strive continuously to improve the effectiveness of methods for informing prospective research volunteers about experimental studies, thereby enhancing the protection of their interests. We should test innovations in informed consent in realistic contexts (i.e., in clinical trials) and with randomization, when it is appropriate, at the first opportunity. In this study, we develop a preliminary proposal to improve the quality of informed consent, based on experimentation with informed consent in ongoing clinical trials. We discuss the conceptual, ethical, organizational, and technical bases for such an effort.