It is known that most patients with type I allergy to tree pollens also suffer from intolerance to nuts. To identify allergenic structures common to hazel pollen and hazelnuts, cross-reactivity of patients' IgE was investigated. With use of immunoblotting,.serum IgE from 25 patients displaying type I allergic reactions to tree pollens and intolerance to hazelnuts (group I) bound to the 17 kd major hazel pollen allergen Cor a I (100%) and to the 14 kd hazel pollen profilin (16%). IgE binding to proteins of comparable molecular weights in hazelnut extracts was found (18 kd and 14 kd), suggesting that proteins similar to Cor a I and hazel profilin might be also expressed in hazelnuts. In contrast, only four sera (22%) from 18 patients (group II) with tree pollen allergy but without any case history of nut hypersensitivity showed IgE binding to the 18 kd protein of hazelnut extract, and none of these sera exhibited IgE reactivity to the hazelnut profilin. To characterize the hazel pollen and hazelnut allergens, purified recombinant Bet v I (major birch pollen allergen) and purified recombinant Bet v II (birch profilin), respectively, were used for IgE-inhibition experiments. Binding of IgE from patients (with nut allergy) to the blotted hazelnut allergens could be blocked by preincubation of patients' sera with the recombinant proteins. Furthermore, the 18 kd protein of hazelnut extract was purified and induced specific release of histamine from basophils of a patient suffering nut hypersensitivity but not from a healthy control donor. A rabbit antibody raised against celery profilin identified the 14 kd proteins in hazel pollen and hazelnuts as profilin. Our experiments suggest a protein with IgE binding properties similar to the major allergens from pollens of hazel, Cor a I, and of birch, Bet v I, as predominant allergens in hazelnuts, and show that the plant pan-allergen profilin can be detected in both hazel pollen and hazelnut extracts.