Background Acute kidney injury (AKI) is a major contributor to morbidity and mortality in neonates with hypoxic–ischemic encephalopathy (HIE). This study evaluated the incidence, clinical predictors, and diagnostic agreement of AKI in neonates with HIE undergoing therapeutic hypothermia (TH).Methods This retrospective cohort study included neonates with HIE treated with TH in a tertiary neonatal intensive care unit between January 2015 and September 2025. AKI was defined and staged according to the neonatal Kidney Disease: Improving Global Outcomes (nKDIGO) and Gupta criteria. Demographic, perinatal, clinical, and laboratory data were compared between AKI and non-AKI groups.Results Seventy-nine neonates were included. AKI occurred in 47% and 33% of patients according to the nKDIGO and Gupta criteria, respectively. Agreement between classifications was weak (κ = 0.30), with diagnostic discordance in 34% of cases. Initial lactate was independently associated with nKDIGO-defined AKI (OR = 1.2; P = 0.047). When the Gupta criteria were applied, the initial Thompson score (OR = 1.3; P = 0.02), hyponatremia (OR = 5.3; P = 0.04) and thrombocytopenia (OR = 16.2; P = 0.02) were independently associated with AKI. The multivariable models showed good discriminative ability for both nKDIGO-defined AKI (AUC = 0.84) and Gupta-defined AKI (AUC = 0.90).Conclusions AKI is common in neonates with HIE undergoing TH, although its incidence and associated predictors varied according to the diagnostic criteria applied. The weak agreement between nKDIGO and Gupta suggests that these classifications may capture different patterns of renal dysfunction. These findings support the need for greater standardization of neonatal AKI definitions and reinforce the importance of early renal monitoring in this high-risk population.