OBJECTIVE To test the hypothesis that the prevalence of deletion 22q11.2 among individuals who meet criteria for DiGeorge anomaly (DGA) is lower than the 90% commonly cited. METHODS Participants were identified through retrospective chart reviews on all patients who underwent testing for deletion 22q11.2 and all patients with a diagnosis of "DiGeorge" or any of the major criteria associated with DGA at a large pediatric hospital over a period of 6 years. DGA was confirmed in 64 individuals, based on the presence of at least 2 of the following features: (1) cellular immune deficiency and/or absence of part or all of the thymus; (2) hypocalcemia and/or parathyroid deficiency; (3) congenital heart disease. RESULTS Of the 64 individuals with DGA, 29 (45%) did not have a chromosome 22q11.2 deletion. Among this deletion-negative subset, diabetic embryopathy and other chromosome abnormalities were the most commonly recognized underlying etiologies. CONCLUSIONS These findings challenge a widely held belief that nearly 90% of DGA is due to chromosome 22q11.2 deletion. This study also calls attention to the heterogeneity of DGA, highlights similarities and differences between those with and without a chromosome 22q11.2 deletion, and attempts to resolve some confusing features of conditions associated with DGA.