This paper reports the results of a multicentric randomized clinical trial on the treatment of first hematological relapse in childhood ALL. Induction treatment consisted of vincristine, adriamycin, L-asparaginase, and prednisone. Patients achieving complete remission were randomized to two maintenance regimens (A and B). Regimen A consisted of five different drug associations including VM26 and IDMTX in a sequential schedule; Regimen B was essentially classical Spiers schedule for the first year, followed by a milder treatment. Eighty-four of 102 evaluable patients (82%) achieved second complete remission. The two maintenance regimens were similar as regards duration of second complete remission (median duration A, 32 weeks; B, 37 weeks) and toxicity. Better results were obtained in patients relapsing after 12 months from suspension of treatment in first complete remission than in those relapsing within the first year off therapy (82.8% vs. 31.4%). In group A fewer CNS relapses were reported. The two regimens produced results similar to those reported by other authors. The good prognosis in patients relapsing at least 1 year after treatment suspension in first complete remission must be emphasized.