Abstract

Bioethicists broadly agree that there is a limit to the level of net risk that biomedical research may permissibly impose on participants, even in cases where the potential of that research to improve the health of the population health would be great. Although some may permissibly volunteer to take on some degree of pro‐social risk, no one, not even a willing volunteer, may ever be outright sacrificed for others. One might think this perspective, if correct, makes it effectively impossible to study interventions with high intrinsic risks. But I describe a method – risk dilution – by which an extremely high‐risk intervention can nonetheless be effectively studied in the context of an acceptably low‐risk trial. I then defend risk dilution from objections.