In recent years, it has been possible to demonstrate mediator release into the nasal secretion after nasal allergen challenge in patients with allergic rhinitis. Using the nasal provocation model, we determined whether the mediator release was altered in immunotherapy-treated patients. Seventeen grass-pollen-allergic patients were examined under controlled, reproducible conditions. Serial challenges with increasing doses of grass pollen produced increasing numbers of clinical symptoms and release of mediators such as kinins, TAME-esterase activity, and histamine. Ten patients received a semidepot perennial grass-pollen extract for 4 years. Seven patients served as controls and did not receive immunotherapy during the observation period. Data from the group of patients receiving immunotherapy over the first year already showed a partially significant decline in the maximal mediator release after nasal allergen challenges compared to the results of pretreated challenges, whereas controls did not show any significant changes. Nasal allergen challenges after termination of 4 years' immunotherapy significantly modified the mediator release compared to pretreatment values (TAME-esterase activity P < 0.05, kinins P < 0.01, and histamine P < 0.01).Decrease of mediator release paralleled the symptom-medication scores and quantitative skin prick test. Finally, we could demonstrate a significant correlation between specific IgG increase and mediator decrease in the treated group.Immunotherapy in patients with pollen allergy has been evaluated by the following criteria: 1) clinical parameters (symptom/medication scores) 2) rhinomanometry (difficult to assess because of 3) direct inquiry, which is highly subjective on the 4) the quantitative skin prick test (QST).Better reproducibility of results was achieved with immunoglobulin measurements (specific IgG and specific IgE), although a statistically significant relation has so far been detectable only between the applied cumulative dose during immunotherapy and the production of IgG antibodies. A direct correlation has not yet been clearly demonstrated between the clinical effectiveness of immunotherapy and IgG antibody production poor reproducibility) part of both the physician and the patient