BACKGROUND: Childhood food allergy is a serious health problem. However, little is known about the frequency and manner in which it is currently diagnosed. OBJECTIVE: To describe parent report of physician practices in the diagnosis of pediatric food allergy. METHODS: Data from children with food allergy were identified for analysis from a representative survey administered in US households with children from June 2009 to February 2010. Analyses were performed at the level of the allergy. Demographic characteristics, symptom prevalence, and diagnostic methods were calculated as weighted proportions. Adjusted models were estimated to examine the association of reaction history and allergenic food with odds of physician diagnosis and testing. RESULTS: Food allergies (n = 3,218) to 9 common allergens were reported among 2,355 children in a sample of 38,480. We found that 70.4% of reported food allergy was diagnosed by a physician. Among physician-diagnosed food allergy, 32.6% was not evaluated with diagnostic testing, 47.3% was assessed with a skin prick test, 39.9% with a serum specific IgE test, and 20.2% with an oral food challenge. Odds of physician diagnosis and testing were significantly higher for severe versus mild/moderate food allergy. Urticaria and angioedema were not reported as symptoms in 40.7% and 34.6% of severe food allergies, respectively. CONCLUSION: Thirty percent of parent-reported food allergies in this study were not diagnosed by a physician. One in 5 physician-diagnosed allergies was evaluated with oral food challenge. Understanding parent report of practices in food allergy provides insight into ways in which to streamline the diagnosis and management of care.