Single lots of food allergen extracts from three different commercial sources were compared for their efficacy in evaluating immediate food hypersensitivity. Eighty-seven children with atopic dermatitis and food hypersensitivity underwent prick skin testing to a battery of 18 food extracts from each company. Results of skin tests were compared with results of double-blind, placebo-controlled oral food challenges and open challenges to determine the sensitivity, specificity, and predictive indices of each reagent. Negative predictive indices were generally good for all reagents, whereas positive predictive indices were generally poor and showed considerable variation (0% to 79%) between commercial sources. Under the conditions of the study, skin test reagents from two companies showed slightly better agreement with double-blind, placebo-controlled food challenge results than did reagents from the third company. However, with known lot-to-lot variations in extract potency and intrapatient variation in skin test results, these differences probably are not of clinical significance.