Forty-five patients with classic food-allergic symptoms and/or subjective food-related complaints not traditionally associated with food allergy underwent evaluation. On the basis of a comprehensive clinical history, skin testing, and placebo-controlled, double-blind food challenges, patients were assigned to one of two groups: patients with reactions highly suggestive of IgE-mediated food hypersensitivity (group A, N = 22) and patients with atypical adverse food reactions that could not be confirmed by double-blind food challenge (group B, N = 23). Most patients in both groups were female, 77.3% and 91.3% of patients in group A and B, respectively. In group B, onset of symptoms occurred at an older age than in group A, 28.9 years +/- 17.2 versus 17.1 +/- 12.1 (p = 0.0015), respectively, and involved more foods, 25.6 +/- 22.1 versus 5.2 +/- 5.5 (p = 0.0002). Foods causing most prominent symptoms among patients in group A included legumes, tree nuts, crustaceans, and fish. In group B, milk, white sugar, wheat, egg, smoked/cured meat, and yeast were among the most troublesome foods. All but one patient in group A gave a positive skin test response to food; only four patients in group B had a positive response. We conclude that a subset of patients with food-related complaints can be accurately predicted to have a negative double-blind challenge with suspected foods on the basis of information obtained by history and skin testing.