Allogeneic stem cell transplantation is the only known curative therapy for myelodysplastic syndromes (MDS). We present the transplant outcomes for 84 adult MDS patients, median age 50 (18-69 years), undergoing allogeneic hematopoietic stem cell transplantation (HSCT) at the University of Minnesota between 1995 and 2007. By WHO criteria 35 (42%) had refractory anemia with excess blasts (RAEB-1 or 2), 23 (27%) had refractory cytopenia with multilineage dysplasia (RCMD) or RCMD and ringed sideroblasts (RCMD-RS), and the remaining 26 (31%) had refractory anemia (RA), myelodysplastic syndrome-unclassifiable (MDS-U), chronic myelomonocytic leukemia (CMML), myelodysplastic/myeloproliferative disease (MDS/MPD), or myelodysplastic syndrome-not otherwise specified (MDS-NOS). Graft source was related in 47 (56%), unrelated donor (URD) marrow in 11 (13%), and unrelated cord blood (UCB) in 26 (31%). The conditioning regimen included total body irradiation (TBI) in 94% of transplantations; 52 (62%) myeloablative (MA) and 32 (38%) nonmyeloablative (NMA) regimens. Cumulative incidence of neutrophil engraftment by day +42, acute graft-versus-host disease (aGVHD) by day +100, and chronic GVHD (cGVHD) by 1 year were 88% (80%-96%, 95% confidence interval [CI]), 43% (36%-50%, 95% CI), and 15% (10%-20%, 95% CI), respectively. One-year treatment-related mortality (TRM), relapse, disease-free survival (DFS), and overall survival (OS) were 39% (28%-50%, 95% CI), 23% (12%-32%, 95% CI), 38% (28%-48%, 95% CI), and 48% (38%-58%, 95% CI) respectively. Cumulative incidence of relapse at 1 year in patients with pre-HCT complete remission (CR) or <5% blasts was improved at 18% (8%-28%, 95% CI) compared to 35% (16%-54%, 95% CI) in patients with 5%-20% blasts (P = .07). Additionally, with MA conditioning, the incidence of relapse at 1 year trended lower at 16% (6%-26%, 95% CI) versus 35% (18%-52%, 95% CI) in NMA (P = .06), and a statistically significant decrease in relapse was noted in patients entering HCT with CR or <5% blasts with an incidence of 9% (0%-18%, 95% CI) (MA) versus 31% (11%-51%, 95% CI) (NMA) (P = 0.04). For those patients with > or =5% blasts, MA conditioning did not significantly decrease relapse rates. One-year TRM was similar between MA and NMA conditioning. For patients entering transplant in CR or with <5% blasts, prior treatment to reach this level did not impact rates of relapse or transplant-related mortality when all patients were analyzed; however, when broken down by conditioning intensity, there was a trend toward improved DFS in those NMA patients who were pretreated. Finally, 1-year DFS was similar using related donor peripheral blood stem cell (PBSC)/marrow, URD marrow, or UCB grafts. These data suggest that (1) blast percentage <5% at HSCT is the major predictor of improved DFS and relapse and prior treatment to reach this disease status may have value in leading to improved DFS; (2) MA conditioning is associated with lower relapse risk, particularly in patients with CR or <5% blasts, but is not able to overcome increased disease burden; (3) NMA conditioning yields equivalent TRM, DFS, and OS, and is reasonable in patients unsuited for MA conditioning; (4) the donor sources tested (PBSC, bone marrow [BM], or UCB) yielded similar outcomes.