A study of the functioning of an institutional review committee, when combined with the Barber et al. survey data on 300 institutions, allows for a useful provisional assessment of the strengths and weaknesses of existing procedures for protecting human research subjects. The evidence shows that there is an important role for such committees; however, most committees are relatively permissive. While one committee's performance showed that such committees can have an impact on proposed research, the absence of effective monitoring procedures and lack of feedback from research subjects limits the impact of the review procedure on the actual conduct of research. The question is also raised whether the review procedure, because of its legitimizing functions, may have some regressive consequences for the protection of human subjects.