Objective: This article reviews, discusses, and elaborates considerations and recommendations summarized by the biological research working group at the May 1993 NIMH conference on ethical issues in mental health research on children and adolescents., Method: Notes from the conference were summarized and supplemented by a computer search of relevant literature. Drafts were circulated for comment to national and international experts, some of whom joined as coauthors., Results: Issues addressed include possible overprotection by policy makers and institutional review boards arising out of the recognition of children's special vulnerability without equal recognition of their need for research; the definition of minimal risk, which has often been equated with no risk in the case of children; assessment of the risk-benefit ratio; procedures for minimization of risk, such as improved technology, "piggybacking" onto clinical tests, and age-appropriate preparation; the difficulty of justifying risk for normal controls; age-graded consent; special considerations about neuroimaging; "coercive" inducement, both material and psychological; disposition of unexpected or unwanted knowledge about individuals, including the subject's right not to know and parent's right not to tell; and socioeconomic status and cultural/ethnic equity., Conclusions: The working group adopted a position of advocacy for children's right to research access while recognizing that this advocacy must be tempered by thoughtful protections for child and adolescent subjects.