Forty-three patients with type I allergic respiratory diseases received aqueous grass pollen immunotherapy in a prospective study designed to compare the clinical and biological effects of a semi-rush hyposensitization (20 patients) with ordinary hyposensitization (23 patients). Before the start of the therapy (end of previous grass pollen season, t1) and before (t2) and after the next grass pollen season (t3), total serum IgE (tIgE), grass pollen-specific IgE (GPsIgE), grass pollen-specific IgG (GPsIgG) and complement factors C3 and C4 were determined. GPsIgE and GPsIgG levels increased significantly more during immunotherapy in patients receiving the semi-rush scheme as compared with the ordinary scheme. There were no significant changes between the two schemes at any times for tIgE, C3 and C4. GPsIgG was already higher in allergic patients before the start of therapy as compared with non-allergic patients and controls. Clinical improvement was the same in patients treated with the semi-rush scheme as in the ordinary group. The number of side effects was not statistically different between the two groups. In neither group was there a correlation between the change in biological parameters and change in clinical symptoms. As the semi-rush scheme gives the same results immunologically as well as clinically, it could be used in a number of patients since it reduces the time course of the initial therapy by about 2 months.