We prospectively treated 127 children with ALL with a risk-adapted regimen. All patients received the identical induction-consolidation therapy. The early maintenance included intermediate dose methotrexate in patients with standard risk (n = 79) and medium risk (n = 39). In addition patients with high risk (n = 6) received high dose ARA-C followed by L-asparaginase. Intensification treatment and prophylactic cranial irradiation was also tailored according to the risk group. Treatment duration was 2 years. Complete remission was achieved in 97.6% of all patients. Treatment-related toxicity accounted for one death in complete remission. The probability of event-free survival (pEFS) for the combined group was 72% at a median follow-up of 42 months. The pEFS was higher in patients with standard risk (SR) than in patients with medium risk (MR) (80% versus 65%; p less than 0.05) at 30 months, but attenuated in the follow-up evaluation at 42 months (76% versus 63%; p less than 0.1). The number of high-risk patients was too small for statistical evaluation. Relapse within the first 18 months after diagnosis indicated a poor prognosis and was more common in patients with MR than in patients with SR. The immunophenotype of the leukemic cells was not found to constitute an independent risk factor after treatment has been risk-adapted. Patients with an initial white blood cell count of more than 50 X 10(9)/l had a worse prognosis than patients with a lower white blood cell count (p less than 0.01).