Compared the behavior of children in 21 child-abusive and 21 nonabusive matched comparison families, using home observations performed by professional independent observers and parental report measures (Becker Bipolar Adjective Checklist, Child Behavior Checklist, and Parent Daily Report). Parental report measures of child conduct problems differed significantly between the groups, with independent observations showing few significant differences in rates of either child or parental behavior. Results are discussed in terms of the validity of parental perceptions and reports of child problem behaviors, the informativeness of differing reports for diagnosis and clinical treatment, and the need to gather corroboration across various sources and settings before clinical decisions are made that may irrevocably alter a child’s future development.