Oral Allergy Syndrome (OAS) in patients with pollen-induced rhinoconjunctivitis is caused by specific IgE recognizing cross-reacting epitopes of fruits and plants, which were clearly shown in vitro, but failed to be demonstrated in vivo by cross-challenges in the target organs. Considering the hypothesis of degradation of such epitopes in natural extracts, challenges with recombinant pollen allergens were done to evaluate the reactivity of the oral mucosa in OAS patients. Seventeen patients with OAS and rhinitis from birch (10) and grass pollen (7) and 10 non-atopic controls were studied by skin prick tests (SPT), allergen specific nasal challenges (ASNC) and allergen specific sublingual challenges (ASSC) with birch and timothy extracts and with rBet v1 and rPhl p1 at increasing concentrations from 1 to 1000 mcg/ml. None of the healthy subjects in the control group had any positive test for birch and timothy extracts or for recombinant allergens. In the OAS group the following results were observed: SPTs with recombinant allergens were positive in all patients, mostly at 10 mcg/ml concentration; ASNC with rBet v1 were positive in all patients, mostly at 100 mcg/ml; ASSC with natural pollen extracts were positive in only 2 of 17 patients, but in 15 of 17 with rBet v1 and rPhl p1, mostly at 500 mcg/ml and 1000 mcg/ml. ASSC with rBet v1 and rPhl p1 were positive with a mean concentration of 677 and 533 mcg/ml, respectively. The results of sublingual challenges with rBet v1 and rPhl p1 showed the in vivo cross-reactivity between pollens and foods in patients with OAS, but high concentrations of the recombinant allergens were needed to reproduce oral symptoms, thus explaining the failure of challenges performed with natural extracts, which have concentrations of major allergens lower than 50 mcg/ml. This indicates that sublingual mucosa is much less reactive to allergens than other surfaces, such as skin and nasal mucosa, probably because of its anatomic and immunologic peculiarity.