The ethical appropriateness of clinical research depends on protecting participants from excessive risks. Yet no systematic framework has been developed to assess research risks, and as a result, investigators, funders, and review boards rely only on their intuitive judgments. Because intuitive judgments of risk are subject to well-documented cognitive biases, this approach raises concern that research participants are not being adequately protected. To address this situation, we delineate a method called the systematic evaluation of research risks (SERR), which evaluates the risks of research interventions by comparing these interventions with the risks of comparator activities that have been deemed acceptable. This method involves a 4-step process: (1) identify the potential harms posed by the proposed research intervention; (2) categorize the magnitude of the potential harms into 1 of 7 harm levels on a harm scale; (3) quantify or estimate the likelihood of each potential harm; and (4) compare the likelihood of each potential harm from the research intervention with the likelihood of harms of the same magnitude occurring as a result of an appropriate comparator activity. By explicitly delineating, quantifying, and comparing the risks of research interventions with the risks posed by appropriate comparator activities, SERR offers a way to minimize the influence of cognitive biases on the evaluation of research risks and thereby better protect research participants from excessive risks.