To explore the feasibility and potential advantages of PBSC in allogeneic transplantation, we grafted 24 patients (age 16-57, median 37) with different hematologic diseases (ALL = 10, AML = 5, MM = 4, NHL = 2, CML = 1, MDS = 1, AA = 1), 23 HLA-identical to their siblings and 1 partially matched. Cells were collected from donors by apheresis after G-CSF 10 to 16 mg/kg/day for 4 to 5 days, and stored at 4 degrees C until infusion. The patients were conditioned with chemotherapy regimens including busulfan and cyclophosphamide in the majority of cases and received GVHD prophylaxis with CSA-MTX in all but two. The graft consisted of PBSC alone, with a median of 143.5 (range 18.1-358.9) x 10(4)/kg CFU-GM, 9.0 (range 3.3-18.0) x 10(6)/kg CD34+ cells and 2.8 (range 1.2 to 8.6) x 10(8)/kg CD3+ and cells. An ANC >0.0.5 x 10(9)/L was recovered on (median) day 13 (range 11-17), and a platelet count >50 x 10(9)/L on (median) day 13 (range 12-55) post graft. There was no correlation between CD34+ cells or CFU-GM number in the inoculum and time to hematologic reconstitution. Acute GVHD (grade II-IV) occurred in 10 out of 22 (45%), chronic GVHD in 10 out of 18 evaluable (55%) patients. We found no relationship between occurrence of acute or chronic GVHD and number of CD3+ cells in the graft. Four patients relapsed and 7 died after transplantation. Fifteen patients are currently alive and disease-free 67 to 710 (median 286) days from the graft. Allogeneic transplantation with unmanipulated PBSC ensures a fast and stable engraftment. Acute GVHD incidence and severity seems comparable to that of bone marrow transplantation, but there may be an increase in chronic GVHD, mainly of the extensive form.