Delayed neutrophil recovery and higher incidence of infections episodes have been observed after unrelated cord blood transplant (UCBT). However, no data is available on the incidence, type and risk factors for severe infections after UCBT. We have analyzed retrospectively 510 UCBT performed from 1994 to 2002 in 55 Eurocord centers. The median age was 6.9 years, the median follow-up for survivors was 36.6 months (3–100); 396 (77%) patients had hematological malignancies, 65 (13%) inborn errors and 49 (10%) bone marrow failure. Conditioning was TBI based in 50% and associated to ATG/ALG in 86%. CsA and corticoides for GVHD prophylaxis was frequently used (77%) and 269 patients (53%) received hematopoietic growth factor. The donor was more frequently 5/6 (n=216, 43%) or 4/6 (n=204, 41%) HLA disparate. The median number of nucleated cells and CD34+ cells infused was 3.8x10e7/kg and 1.6 x10e5/kg, respectively. 95% of the patients were transplanted in a LAF room. Prophylaxis of infections varied among centers. Incidence of first infections were analysed using competing risk analysis; risk factors were studied using cumulative hazard analysis of all infectious episodes occuring during 100 days after UCBT.Results: Cumulative Incidence (CI) of neutrophil recovery, acute GVHD (II–IV), and 100 days mortality were 75%, 38% and 32% respectively. CI at day 100 of first overall, bacterial, viral and fungal infections were respectively 69%, 49%, 32% and 10%. During the first 100 days, 686 severe infections episodes were diagnosed in 352 patients. A total of 404 episodes were from bacteria origin: 276 gram +; 124 gram- and 4 others, 189 severe viral infections or diseases episodes (142 CMV, 21 adenovirus, 12 EBV and 12 Herpes simplex or HHV6 and 2 para-myxo-virus, 54 episodes of severe fungal infections (26 candidemia, 20 aspegillus and 8 other), 5 episodes were of toxoplasmosis and 34 episodes were not classified. In a multivariate analysis the following factors decreased the cumulative hazard of 1) bacterial infections: shorter time to engraftment (HR: 0.22, p<0.0001); 2) viral infections: negative CMV serology (HR: 0.28, p<0.001); less than 3 out of 6 HLA disparate (HR: 0.21, p=0.03) and shorter time to engraftment (HR: 0.40, p=0.002); and finaly for 3) fungal infections: recipient's age <16 years (HR: 0.10, p=0.004), malignant disease (HR:0.13, p=0.03), shorter time to engraftment (HR:0.17, p=0.02); and absence of acute GVHD III-IV (HR 0.24, p=0.03). Considering all types of infections, among the factors cited above, period of UCBT performed (after 1998) was also associated with a decreased hazard of severe infections at day 100.In conclusion, in this retrospective analysis, bacterial infections appeared early, followed by viral and fungal infections during the first 100 days after UCBT. Delayed engraftment was frequently associated with increased risk of all types of infections. Aproaches that will improve time to engraftment after UCBT might decrease the incidence and severity of early infections after UCBT.