OBJECTIVE: The aim of this study was to determine graft function and survival for kidney transplants from deceased donors with acute kidney injury (AKI) that persists at the time of organ procurement. BACKGROUND: Kidneys from donors with AKI are often discarded and may provide an opportunity to selectively expand the donor pool. METHODS: Using Organ Procurement and Transplantation Network and DonorNet data, we studied adult kidney-only recipients between May 1, 2007 and December 31, 2016. DonorNet was used to characterize longitudinal creatinine trends and urine output. Donor AKI was defined using KDIGO guidelines and terminal creatinine ≥1.5 mg/dL. We compared outcomes between AKI kidneys versus "ideal comparator" kidneys from donors with no or resolved AKI stage 1 plus terminal creatinine <1.5mg/dL. We fit proportional hazards models and hierarchical linear regression models for the primary outcomes of all-cause graft failure (ACGF) and 12-month estimated glomerular filtration rate (eGFR), respectively. RESULTS: We identified 7660 donors with persistent AKI (33.2% with AKI stage 3) from whom ≥1 kidney was transplanted. Observed rates of ACGF within 3 years were similar between recipient groups (15.5% in AKI vs 15.1% ideal comparator allografts, P = 0.2). After risk adjustment, ACGF was slightly higher among recipients of AKI kidneys (adjusted hazard ratio 1.05, 95% confidence interval: 1.01-1.09). The mean 12-month eGFR for AKI kidney recipients was lower, but differences were not clinically important (56.6 vs 57.5 mL/min/1.73m 2 for ideal comparator kidneys; P < 0.001). There were 2888 kidneys discarded from donors with AKI, age ≤65 years, without hypertension or diabetes, and terminal creatinine ≤4 mg/dL. CONCLUSION: Kidney allografts from donors with persistent AKI are often discarded, yet those that were transplanted did not have clinically meaningful differences in graft survival and function.