Umbilical cord blood is a useful stem cell source for patients without matched related or unrelated donors. However, single cord blood unit transplantation in adults is associated with high transplant related mortality, mostly due to infection. In this study, we used a reduced intensity conditioning regimen followed by infusion of two partially matched cord blood units. The conditioning regimen was fludarabine 30mg/m2/day x 6 days, melphalan 100mg/m2/day x 1 day, and rabbit antithymocyte globulin 1.5 mg/kg/day x 4 days. Cord blood units were a 4/6 or better HLA match or better with each other and with the patient, and contained a minimum combined pre-freeze cell dose of 3.7 x 107NC/kg. GVHD prophylaxis was cyclosporine and mycophenolate mofetil. Twenty-one patients, 15 males (71%) and 6 females (29%), median age 49 years (range 24–63 years) participated in a Phase I study. The diagnoses were AML (n=8), ALL (n=1), NHL (n=5), CLL (n=2), MDS (n=2), Hodgkins Disease (n=2), and aplastic anemia (n=1). Fifteen percent of patients were non Caucasian. The cell doses infused were a median of 4.0 x107 NC/kg (range 3.0–5.3 x107) and 2.0 X105 CD34+ cells/kg (range 0.6–10.0 x105). Two patients (both with MDS complicating aplastic anemia) experienced primary graft failure, and received successful second cord blood transplants using a different conditioning regimen. Among the remaining 19 patients, the median time to an absolute neutrophil count >500 was 20 days (range 15–34 days). The median time to a platelet count >20,000 unsupported were 41 days (range 21–125 days). One patient experienced a secondary graft failure, and is well following infusion of previously stored autologous cells. 4 patients (21%) experienced Grades II-IV acute GVHD, and only one patient (5%) experienced Grade III GVHD. There were no patients with Grade IV GVHD and no deaths from acute GVHD. Twelve patients were evaluable for chronic GVHD, and 3 patients (25%) had chronic GVHD of which one case was extensive disease. The 100 day transplant related mortality was 14%. The deaths were due to a CNS bleed, Epstein Barr virus lymphoproliferative disorder, and staphylococcal sepsis. Chimerism analysis showed predominance of one cord by Day +100 in 79% of patients evaluable for 100-day follow-up. In 85% of these patients the first cord blood unit infused predominated. One patient has had progressive disease. With a median follow-up of 7 months (range 2–16 months), the overall survival is 79% and the disease-free survival is 64%. The projected one year disease-free survival is 64%. In conclusion, 1) engraftment of adult patients appears to be acceptable using double cord blood products and reduced intensity, non TBI conditioning regimen; 2) the risk of serious acute and chronic GVHD is low, 3) patients with aplastic anemia/MDS may require more intensive immunosuppression to allow engraftment, 4) GVL appears to be preserved despite the low T cell dose.