Canine kidney preservation models have historically used autotransplants to avoid the complications of rejection, although clinically all transplants are allografts. This study investigated the effects of preservation time and method on early kidney function in a canine allograft vs. autograft model. Kidneys were harvested from beagles, preserved by cold storage (CS) in UW solution for 0, 24 or 72 h, or by machine perfusion (MP) with Belzer MPS for 72 h. In some experiments 45 min of warm ischemia (WI) was performed in situ before harvest. Allograft recipients received steroid immunosuppression. Kidney function was assessed by serum creatinine and survival for 7 days. Allografts preserved for 0 and 24 h performed as well as autografts. Allografts preserved for 72 h by either CS or MP had a higher incidence of primary nonfunction (PNF) compared with autografts, as determined by survival (50% vs. 100%, p < 0.003). Primary nonfunction kidneys had thrombotic microangiopathy, vascular and peritubular capillary binding of IgM and complement C4d, and evidence of circulating donor-specific antibodies; all consistent with humoral rejection. These responses were dependent on hypothermia time and were not attributable to ischemia, immunosuppression, preservation solution, or cellular rejection. In conclusion, prolonged hypothermia can cause PNF in allografts owing to acute humoral rejection.