A retrospective study was performed on 46 unselected acute lymphoblastic leukaemia (ALL) elderly patients aged 60 years or more. Only 50% of these patients were included in the EORTC cooperative clinical trials, thus confirming the important selection bias in most of the published series on elderly ALL patients. 43% of the elderly patients achieved a complete remission (CR). The median survival was 10 months and the 5-year overall survival was only 7.6 +/- 4%. In multivariate analysis, W.H.O. performance status and peripheral blast counts at day 7 were found to significantly influence achievement of CR and survival. In patients with W.H.O. performance status > or = 2, 35% died during induction treatment versus 4% in patients with W.H.O. performance status < 2. Patients > 70 years old showed a marked drop of the CR rate (27%) compared to those aged 60-69 (67%), and a very high death rate during the induction period (38% versus 4%). This suggests that ALL protocol treatments should be proposed until 70 years in patients with good-performance status, whereas less intensive treatment should be offered to elderly patients with performance status > or = 2 and/or age > or = 70. Peripheral blast counts at day 7 may help to adjust the treatment during induction phase.