The diagnosis of type I allergy, an IgE-antibody-mediated hypersensitivity disease affecting more than 25% of the population, is based on the measurement of allergen-specific serum IgE levels and provocation testing. Whether the determination of allergen- specific serum IgE levels can replace in vivo provocation testing for allergy diagnosis is a controversial issue. We used purified recombinant timothy grass and birch pollen allergens to compare by skin prick and nasal provocation testing as well as by serology in vivo sensitivity with antibody-binding capacity in 24 pollen allergic patients and eight control individuals. Results from biologic tests were correlated with each other and with allergen-specific IgE and IgG1-4 levels. IgE-reactive allergens induced immediate skin and nasal reactions, but the intensity of the allergic tissue reactions was not correlated with either the levels of allergen-specific IgE or the levels of allergen-specific IgG antibodies. Less frequently detected allergens with low IgE-binding capacity were able to induce strong allergic reactions comparable to those caused by major allergens with high IgE-binding capacity. In contrast, skin test and nasal provocation results were significantly correlated (r = 0.63, p < 0.01). Our study thus demonstrates on a molecular level that skin testing provides a better reflection of immediate type respiratory sensitivity than serologic measurements. These results have implications for allergy diagnosis and, in particular, for the selection of relevant allergen components for specific immunotherapy.