Exome sequencing is the current standard for diagnosing Mendelian disorders
however, it is generally not considered the first-line test for detecting copy number variants (CNVs). We retrospectively investigated the additional diagnostic yield by performing concurrent CNV analysis using exome data in a large and diverse pediatric cohort. Patients were referred from various sources with variable phenotypes. Human Phenotype Ontology terms were used to prioritize variants for analysis. Ancestry and CNV analyses were performed using Somalier and NxClinical, respectively. A total of 1538 patients were tested, with the majority being Admixed Americans. Diagnostic CNVs were identified in 70 patients (4.6%), ranging from exonic deletions to large, unbalanced rearrangements, aneuploidies, and mosaic findings. While no significant differences were identified in diagnostic yield, or rates of negative or uncertain diagnoses, between ancestries, our study demonstrates the feasibility and increased yield of CNV analysis of exome data, across multiple phenotypes, referral sources, and ancestries.