Background. Risk factors for delayed graft function (DGF) in pancreas transplantation (PTx) and its implications on graft survival are poorly defined. Methods. Eighty-seven consecutive first-time PTx for type I diabetes performed between January 2003 and December 2007 were retrospectively reviewed. DGF was defined as a reversible need for exogenous insulin beyond postoperative day 10 (DGF group [DGFG]). For statistical analysis, DGFG patients were compared with patients with immediate graft function (control group [CG]). Results. DGF occurred in 16 patients (18.6%). C-peptide levels and DGF were inversely correlated (r = 0.24, P = 0.03). In univariate analysis, donor cytomegalovirus (CMV) + antibody status, and D+/R - CMV mismatch were significantly associated with DGF (81.3% vs. CG 52.1%, P = 0.029; and 62.5% vs. CG 21.1%, P = 0.002, respectively). Compared with University of Wisconsin solution, histidine tryptophan ketoglutarate-preserved grafts displayed higher DGF rates (37.5% vs. CG 12.7%, P = 0.030), similar to female recipients (DGFG 68.8% vs. CG 35.2%, P = 0.015). On multi-variate analysis, a significantly higher DGF incidence was noted in female recipients (DGFG 68.8% vs. CG 35.2%; P = 0.03) and in recipients with D+/R - CMV mismatch (DGFG 62.5% vs. CG 21.1%; P = 0.03). With a median follow-up of 40.4 months (range 0.7-74.2), graft survival at 5 years did not differ between both groups (94.4% CG vs. 93.8% DGFG; P = 0.791). Conclusion. This is the first study that identifies CMV mismatch (D+/R-) as an additional risk factor for DGF occurrence in PTx. In this particular cohort, DGF does not seem to affect graft survival.