Title: Pediatrician interview style and detection of child abuse. Objective: Generally, to improve the recognition of and response to child maltreatment in the context of primary care pediatrics. Specifically, to test the relationship between a pediatrician's interview style and skills and his/her ability to recognize and respond to cases in which maltreatment may be suspected. Design: A cross-sectional and retrospective study of the relationship between interview skills, referrals of cases in which maltreatment is suspected, and responses to hypothetical cases of suspected abuse. Setting: The pediatric residency program of a large, inner city hospital; specifically, the pediatric primary care clinic. Subjects: Physician subjects: PGY3 residents (n=about 60 in three cohorts of 20). Patients/parents: families who are part of each physician's panel of continuity patients (10 per resident). Measurement/methods: Outcome measures: a) response to clinical vignettes of cases in which maltreatment might be suspected; b) retrospective ascertainment of each physician subject's referrals to social work for suspected maltreatment (number, type of case) during their PGY1 and 2 years. Independent variables will include: a) audiotape analysis of physician-parent interviews; b) physician sensitivity to parent/child psychosocial problems, measured by comparison of physician ratings with parent assessments; c) physician attitudes toward child care standards; d) physician perception of the availability and quality of mental health resources. Significance: The study hope to provide background for improving physician recognition and treatment of child abuse in a way that circumvents other more value-laden and narrowly focused approaches. It will also provide new information on the content of the parent-physician interaction in pediatrics and characteristics that contribute to more effective management of psychosocial problems.