Ten CVID patients with defective IL-2 synthesis in vitro were treated with nhuIL-2 in a placebo-controlled, double blind, crossover therapy study during a period of 12 months. No severe side-effects of nhuIL-2 were recorded. Marginal serum nhuIL-2 levels were measurable in individual patients only during the therapy phase. Serum levels of soluble IL-2 receptors were unaffected by the therapy. nhuIL-2 and placebo groups did not differ significantly with respect to requirement of IVIG substitutions which were performed whenever serum IgG levels dropped below 5 g/l: a total of 53 IVIG infusions (corresponding to 17.6 g IgG/month per patient) was necessary during the placebo phase, and 48 infusions (16.4 g IgG/month per patient) during the nhuIL-2 treatment phase. Thus, nhuIL-2 therapy was ineffective in improving spontaneous IgG synthesis in vivo. Nevertheless, the group of patients receiving nhuIL-2 during the first 6 months of the study exhibited a significant reduction of severe infections (n = 25) during the following 6 months of placebo treatment (n = 7) (P<0.045). The infection score dropped in this group from 181 to 23 (P<0.015). Patients of the second group receiving first placebo and then nhuIL-2 did not experience a significant difference in number and score of infectious episodes: 25 infections were recorded during the first 6 months and 24 during the following 6 months. We suppose that nhuIL-2 therapy of CVID patients reduces susceptibility to severe infections, possibly via the induction of a specific antibody response, which is effective at the earliest 6 months after initiating nhuIL-2 therapy.