Findings suggested that most key variables were intercorrelated, confirming the need to account for confounding among children exposed to complex diversity. Only 20 percent of the SBP dichotomous score variance was accounted for in the two-step regression model, and most associated variables were not independent predictors. Children in family-based alternate care presented with complex mental health difficulties that were poorly conceptualized, and which were not adequately explained by comorbidity; sexual behavior problems represented one facet of this complexity. Almost all children in the study with SBP had corresponding psychopathology, most notably conduct problems, inattention, and interpersonal behavior problems suggestive of attachment disturbances. Evidence of sibling concordance for SBP was found among sibling pairs, without corresponding concordance of sexual abuse. Female gender also predicted in SBP scores independent of girls' higher exposure to sexual abuse. Although an age effect was detected, it was perhaps illusory, confounded by children's ages and entry into care. The study simultaneously examined a large number of discrete and cumulative influences on the development of children at high risk for SBP. The design helped delineate between proximal and distal influences. After controlling for other influences, child sexual abuse (CSA) predicted SBP. The findings also emphasize the significance of cumulative exposure to adversity among children who have been chronically maltreated, and generated several hypothesized attachment mechanisms. Data were collected from caregiver-reported SBP of 4- to 11-year-old children measured in the baseline survey 2000 through 2003 of the Children in Care Study (CICS), a prospective study of the mental health of children in court-ordered care in New South Wales, Australia. Tables, references