Some older patients (> or =40 years) with chronic myelogenous leukemia (CML) who lack human leukocyte antigen (HLA)-identical sibling donors are not offered unrelated marrow transplantation because of concerns over excessive regimen-related toxicity, in particular due to graft-vs.-host disease (GVHD). The purpose of this study was to determine the efficacy and toxicity of unrelated marrow transplantation in older CML patients using a regimen designed to minimize the severity of GVHD. Thirty-one consecutive patients over the age of 40 with CML received unrelated marrow transplants between January 1988 and June 1997. Twenty-one patients were transplanted in chronic phase while ten were transplanted in the accelerated phase of their disease. Fifteen patients received transplants from phenotypically matched donors while 16 received marrow grafts from donors who were mismatched at one HLA locus. GVHD prophylaxis consisted of ex vivo T cell depletion of the donor marrow graft plus posttransplant cyclosporine administration. Durable engraftment was achieved in 29 of 31 patients (94%). The probability of developing grades II-IV or severe grades III-IV acute GVHD was 39.2 and 7.1%, respectively. There was no difference in the incidence of grades II-IV acute GVHD between patients transplanted with marrow grafts from phenotypically matched (38.1%) vs. those transplanted from mismatched unrelated donors (40%, p = 0.99). The 2-year probability of relapse for the entire population was 29.4%. Relapse was significantly higher for patients transplanted in accelerated phase (60%) than for those in chronic phase (13.8%, p = 0.027). The 2-year probability of overall survival and disease-free survival for the entire cohort was 56 and 45%, respectively. There was no significant difference in survival or disease-free survival for patients receiving phenotypically matched vs. mismatched marrow grafts. Immunological reconstitution for this cohort was compared with a younger (<40 years) patient population that had been similarly transplanted over the same time period. Immune function as assessed by total T cell, B cell, NK cell, and T cell subset reconstitution posttransplant was quantitatively equivalent in the two groups with most parameters normalizing within 18 months of transplant. We conclude that CML patients over the age of 40 who have either phenotypically matched or one antigen-mismatched unrelated donors can successfully undergo allogeneic marrow transplantation. T cell depletion of the marrow graft may be advantageous in these older patients by reducing GVHD severity, particularly in those patients transplanted with HLA-disparate marrow grafts.