Objectives: To validate a three-factor model for the Pediatric Symptom Checklist-17 (PSC-17) and evaluate its diagnostic accuracy with African-American and Caucasian children with and without a chronic illness. Methods: Mothers of 723 youth diagnosed with either type I diabetes (n=210) or sickle cell disease (n=191) and a nonill peer group (n=322) completed a demographic questionnaire, the PSC-17, and the Child Behavior Checklist (CBCL). Results: Confirmatory factor analyses and tests of measurement invariance validated a three-factor structure for the PSC-17 with African-American and Caucasian youth with and without a chronic illness. Receiver operating characteristic curves revealed optimal cut-off scores that are similar to published reports. Conclusions: A three-factor solution was replicated for the PSC-17 with African-American and Caucasian children with and without a chronic illness. Cut-off scores for identifying children at risk for emotional/behavioral problems were evaluated using the CBCL as the gold standard and are discussed. © 2011 The Author. Published by Oxford University Press on behalf of the Society of Pediatric Psychology. All rights reserved.