To be a functional alternative to government regulation, self-regulation of science must be credible to both scientists and the public, accountable, ethical, and effective. According to some, serious problems continue in research ethics in the United States despite a rich history of proposed self-regulatory standards and oversight devices. Successful efforts at self-regulation in stem cell research contrast with unsuccessful efforts in research ethics, particularly conflicts of interest. Part of the cause for a lack of success in self-regulation is fragmented, disconnected (...) oversight, and failure to embody genuine scientific and public consensus. To be accountable, credible and effective, self-regulation must be inclusive and multidisciplinary, publicly engaged, sufficiently disinterested, operationally integrated with institutional goals, and must implement a genuine consensus among scientists and the public. The mechanisms of self-regulation must be sufficiently broad in their oversight, and interconnected with other institutional forces and actors, that they do not create fragmented solutions. (shrink)