Accurate and consistent risk assessment is vital for the protection of research participants. Yet, current evaluation of the risks of research interventions often does not take into account the relevant empirical data. This approach raises concern that current practice may not be protecting research participants adequately, or that it may be thwarting acceptable research.