To compare the efficacy of immunoglobulin replacement therapy given intravenously versus subcutaneously to prevent infections in patients with primary antibody deficiency syndromes, an international, multicenter, open label, crossover study was designed. Forty patients were randomized to receive either subcutaneous or intravenous immunoglobulin replacement therapy for 1 year. In the second year, patients were switched to the alternative treatment, enabling patients to act as their own controls. Equivalent doses were given by both routes. Ethical approval was obtained from the review boards of the hospitals in which the patients were seen and written consent obtained from each patient. Patients with a primary antibody deficiency syndrome, either common variable immunodeficiency or IgG subclass deficiency or specific antibody deficiency, who required immunoglobulin replacement therapy were included in the study. Patients were excluded if they had significant thrombocytopenia (defined as platelets less than 50 x 10(9)/liter), had high levels of anti-IgA antibodies (defined as greater than 1:8192), or had severe adverse reactions to a blood product within the last 2 years. The primary end point was the number of infections and their severity (moderate and major) during the two treatment periods. Secondary end points were adverse reactions, length of infections, days lost from school or work due to infections, and acceptability of treatment regimens to the patients. Based on the assumption that it was difficult to prove equivalence of therapies statistically in crossover studies, an arbitrary number of 40 patients was selected on the basis that this might be achievable in 2 years. There are no significant differences in efficacy or adverse reaction rates between immunoglobulin replacement therapy given subcutaneously or intravenously.