In a randomised study the efficacy of a cytomegalic hyperimmune globulin preparation (CMV-HIGP) which had been treated with beta-propiolactone was analysed. The study included 85 patients with acute lymphoblastic leukemia (ALL) and Non-B-Non-Hodgkin-lymphoma (NHL) who were treated initially or underwent a relapse therapy. During the intense chemotherapeutical period within leukemia treatment the patients were passively immunised by the intravenous route with CMV-HIGP (1 ml per kilogram of body weight) every two to three weeks at the latest. In the initial stages the basic immunisation protection was achieved by the application of double dose CMV-HIGP. The Frankfurt patients were recruited from the BFM-ALL- and the NHL-study since october 1982. When they were admitted their CMV serostatus was determined by means of the ELA-ELISA or IFA-method. Seronegative patients were given the passive immunisation immediately or 48 hours after the first blood transfusions at the latest. The patients who had become CMV-IgG-positive by passive immunisation were randomised when reaching long-term therapy according to the protocol. Because of a 30% cytomegaly disease incidence rate in our patient population a randomisation was unwarrantable at the beginning of leukemia treatment. During randomisation one group of patients were immunised by the intravenous route with CMV-HIGP (2 ml per kg body weight one time in four weeks), the second group was a control group.(ABSTRACT TRUNCATED AT 250 WORDS)