A systematic review was undertaken of scientifically rigorous studies of family-based services in children's health and mental health. From a pool of over 4000 articles since 1980 in health and mental health that examined either specific family-based interventions for families of children or the processes of involvement, 41 studies were identified that met the methodological criteria for inclusion. These 41 studies encompassed 3 distinct categories: (a) families as recipients of interventions (e.g., family education, support, engagement, empowerment); (b) families as co-therapists; and (c) studies of the processes of involvement (e.g., therapeutic alliance, engagement, empowerment, expectancies, and choice). Too few experimental studies exist to conclude decisively that family-based services improve youth clinical outcomes. However, those studies that have been rigorously examined demonstrate unequivocal improvements in other types of outcomes, such as retention in services, knowledge about mental health issues, self-efficacy, and improved family interactions--all outcomes that are essential ingredients of quality care. Four implications are drawn from this review. (1) Effective family education and support interventions from studies of adults with mental illnesses and from studies of families of high-risk infants exist and can be imported into the field of children's mental health. (2) The range of outcomes that are typically assessed in clinical treatment studies is too narrow to afford an adequate view of the impact of family-based interventions. A broader view of outcomes is needed. (3) The absence of a robust literature on process variables other than therapeutic alliance limits conclusions about how and why interventions are effective. Attention to the processes by which families become involved in services will require a more robust and nuanced range of studies that attend simultaneously to processes of change and to outcome improvement. (4) Linkage of effective family-based interventions to delivery of evidence-based services is likely to amplify the impact of those services and improve outcomes for youth and families.