In 1981 the BFM group introduced a new treatment strategy for B-ALL based on two alternating 5-day courses of chemotherapy delivered in short intervals up to a total of eight. The therapy courses were composed of fractionated cyclophosphamide, MTX 0.5 g/m2 (24-h infusion), i.t. MTX therapy, and ara-C/VM26 alternating with doxorubicin. The development of the therapy strategies during the subsequent two studies was characterized by shortening treatment duration from eight to six courses, and intensification of CNS chemotherapy in study ALL-BFM-83, and the introduction of HD-MTX (5 g/m2, 24-h infusion) in study ALL-BFM-86. In study ALL-BFM-81, CNS therapy consisted of ID-MTX in combination with i.t. MTX and RX. CNS-positive patients received complete neuroaxis irradiation. In study ALL-BFM-83, CNS chemotherapy was intensified by adding dexamethasone, while MTX/ara-C were administered intraventricularly. Spinal irradiation for CNS-positive patients was omitted. In study ALL-BFM-86, i.t. MTX/ara-C/prednisolone therapy was introduced in combination with HD-MTX but the intraventricular route of drug administration was no longer used. Radiotherapy was omitted completely. In all, 87 patients were enrolled, 22 (eight CNS positive) in study ALL-BFM-81, 24 (seven CNS positive) in study ALL-BFM-83, and 41 (none CNS positive) in study ALL-BFM-86. The estimated 5-year duration of EFS was 40% in study ALL-BFM-81, 50% in study ALL-BFM-83, and 78% in study ALL-BFM-86 (minimal follow-up 36 months). Nineteen of 24 relapses occurred while on therapy or shortly thereafter. In study ALL-BFM-81, the CNS was the most frequent site of failure. In ALL-BFM-83 there were no isolated CNS relapses but more BM relapses occurred. In ALL-BFM-86 localized manifestations were the predominant site of failure, no isolated BM relapses occurred, and only one CNS relapse was diagnosed. No single parameter exerted a consistent influence on outcome with one exception. The presence of residual disease after the first two courses was correlated with an increased risk of therapy failure. Our conclusions from the three studies are listed below. An intensive, short-pulse therapy delivered within a 4 month period is highly effective in the treatment of B-ALL. Prolonged therapy duration is of no value. In addition to fractionated cyclophosphamide-ifosfamide, a 24-h infusion of HD-MTX5 g/m2 in conjunction with an i.t. therapy is a very important component for prevention of both systemic and CNS relapses.(ABSTRACT TRUNCATED AT 400 WORDS)