Objective: This study examines whether the Child Behavior Checklist (CBCL) can be used as an accurate psychiatric screening tool for children in community mental health settings. Method: Associations, logistic regression models, and receiver operating characteristic (ROC) analysis were used to test the predictive relationship between the CBCL and structured interview-derived diagnoses. Results: Associations between CBCL subscales and specific diagnoses were found, replicating results of previous research. Multivariate models and ROC scoring led to correct diagnostic predictions in a large majority of participants. Corresponding sensitivities, however, were deemed inadequate as they indicated that the screen would miss an unacceptably high number of cases. Conclusions: Using the CBCL as a diagnostic screening tool would result in a large number of cases being missed by the screen. Although the CBCL is unlikely to accurately identify children with specific diagnoses, it could be used as a triage tool to inform practitioners if an in-depth structured diagnostic interview is necessary. (Contains 5 tables.)