Peripheral blood stem cell transplants (PBSCT) from unrelated donors (n = 37) were compared with bone marrow transplants (BM, bone marrow group, n = 37) in a matched pair analysis. Ten patients (2, class 1) in the alloPBSCT group and seven patients (2, class 1) in the BM group had one HLA locus mismatch donor, respectively. The following factors were matched: HLA-compatibility, diagnosis, disease stage, age and gender. The median age in the PBSC group was 37 years (19-56, excluding one 6-year-old child) and in the BM group 37 years (18-53). The BM group consisted of 12 females and 25 males, 17 females and 20 males were in the PBSC group. Twelve patients in the BM and 11 patients in the PBSC group were diagnosed with AMI,; 7/7, ALL; 15/15, CML; 2/3, MDS; 1/1, NHL. Thirty-four (14/20) of the 74 patients (45%) were considered as high risk patients. The conditioning regimen was BU/CY for standard risk patients with myeloid diseases (31 patients) and TBI/CY for ALL and NHL patients (36 patients); six patients received intensified conditioning with VP16 (2 patients), thiotepa (2 patients) or melphalan (1 patient). The GVHD prophylaxis regimen was used according to the Seattle protocol. DFS was 51% (19 patients) with a median of 352 days and 59% (21 patients) with a median of 760 days, in PBSC and BM transplants, respectively. The median time to leukocyte engraftment in PBSC patients was 14 days (range 6-26 days) and in the BM group 19 days (range 9-29 days; P < 0.02). The time of platelet engraftment did not differ significantly between the groups. The incidence of grade II-IV acute GVHD was 40% (four patients died, 13%) in the PBSC group and 20)% (three patients died, 8%) in the BM group, respectively (P < 0.05, log-rank). No signs of aGVHD were found in 19% of the patients in the PBSC and 27% in the BM group. Our results indicate that allogeneic PBSCT does lead to a significantly faster leukocyte engraftment. The significant increase with regard to the incidence and shorter time of onset of severe aGVHD in PBSC patients, compared to marrow transplant patients, need to be confirmed in a randomised trial.