A frequent change of non-crossresistant drug combinations might obviate the problem of multiple resistant cell lines in malignant diseases and thus increase cure rates. In a multicenter cooperative study for childhood acute lymphoblastic leukemia (ALL) 109 high-risk patients were randomized to receive 5-6 different drug combinations either in slow rotation (change of drugs every 4-6 weeks) or in rapid rotation (change of drugs every 2-3 weeks) for early intensive treatment. Both groups received the same total amount of drugs within the same period of time. 108/109 patients achieved complete remission. One child failed to enter remission and one was lost in remission due to viral infection. Patients in the rapid rotation arm required 2-3 weeks less time to complete the intensive therapy due to fewer episodes of prolonged myelosuppression. Toxic side effects and infectious complications were comparable in both groups. 9/109 patients relapsed, 6 in the bone marrow and 3 in the central nervous system. As yet none of the 31 patients with an initial white blood count of greater than or equal to 100/nl has relapsed. The probability of relapse-free survival is 88% in the rapid rotation arm and 86% in the slow rotation arm at 2 1/2 years. The results compare favourably with other protocols for high-risk patients but have still to be considered as preliminary because of the short median observation time of 18 months.