The study was designed (a) to provide prevalence data on behavioral problems and competencies, (b) to identify differences related to demographic variables, and (c) to compare clinically referred and demographically similar nonreferred children. Data were obtained with the Child Behavior Checklist (CBCL), consisting of 20 social competence items and 118 behavior problems. Parents of 1,300 referred children completed the CBCL at intake into outpatient mental health services, while parents of 1,300 randomly selected nonreferred children completed the CBCL in a home interview survey. Intraclass correlations were in the .90s for interparent agreement, 1-week test-retest reliability, and inter-interviewer reliability. Indices of the reported prevalence of each item are graphically portrayed for children grouped by age, gender, and clinical status. Multiple regressions and ANCOVAs showed minimal racial differences but significant tendencies for lower SES children to have higher behavior problem and lower competence scores than upper SES children. There were numerous gender differences on specific items but no significant gender difference in total behavior problem or competence score. Age showed more and larger differences than the other demographic variables, but these differences were dwarfed by differences related to referral status. Across all demographic groups, referal status accounted for more variance in total behavior problem and social competence scores than in the scores for any single item. However, some behavior problems that have traditionally received little attention were much more strongly associated with referral status than problems that have received much attention. Cutoff points on the distributions of total behavior problem and social competence scores yield good separation between referred and nonreferred samples.