This multicenter study was designed to evaluate whether allo-PBPCT provides some advantages, if any, over BMT in terms of engraftment kinetics, acute and chronic GVHD incidence, TRM, relapse incidence and survival in acute lymphoblastic leukemia patients (ALL). From January 1995 to December 1999, 67 ALL patients (34 in the PBPCT group and 33 in the BMT group) were included in this study. Median age for both groups was 8 years (range, 1-18). There were 24 patients in first or second CR in the PBPCT group and 28 such patients in the BMT group. Preparatory regimens were TBI-based in 26/34 in the PBPC group and 25/33 in the BMT group. GVHD prophylaxis was CsA alone in 38 patients (18 PBPCT vs 20 BMT) and CsA plus short Mtx in 29 (16 PBPCT vs 13 BMT). Engraftment was achieved in all cases. Median days to neutrophil recovery was 10 (range, 7-18) after PBPCT vs 14 (range, 9-21) after BMT (P < 0.0001). Platelet engraftment (>50 x 10(9)/l) was also faster for PBPCT patients (median 13 days, range, 9-40 vs 23 days, range, 15-165) (P < 0.0001). Acute GVHD grade II-IV incidence was similar in both groups (46.4 +/- 8.8% vs 42.7 +/- 8.6%) (P = 0.45). Probability of chronic GVHD was 50.6 +/- 12.2% after PBPCT vs 27.8 +/- 9.2% after BMT (P = 0.1). Probability of relapse was similar (28.7 +/- 9.2% for PBPCT vs 27.1 +/- 8.2% for BMT) (P = 0.89). There were eight patients who died from transplant-related complications after PBPCT vs 5 after BMT (P, NS). With a median follow-up of 25 months the event-free survival probability was 53 +/- 8.9% for PBPCT vs 54.9 +/- 9.7% for BMT (P = 0.54). Using PBPC for allogeneic transplantation in childhood ALL results in faster hematopoietic recovery compared to BM, with a similar incidence of aGVHD, TRM, relapse and disease-free survival. However, the issue of cGVHD remains unresolved.