Acute kidney injury (AKI) is a common and increases risk of chronic kidney disease (CKD). While mitochondrial dysfunction drives maladaptive repair, the role of glycolysis in renal recovery remains unclear. Here, we integrated single-cell transcriptomic data from mouse models with clinical transcriptomic and functional data from kidney transplant recipients to investigate how glycolysis influences epithelial repair. Trajectory inference and metabolic flux estimation revealed divergent cellular fates: cell with active glycolysis achieved recovery, while those with blunted glycolysis followed, maladaptive paths. These results were confirmed in an independent single-cell dataset and validated clinically in reperfusion biopsies. A machine-learning model trained on clinical parameters identified patients with underactive glycolysis at reperfusion; this subgroup showed impaired long-term recovery. Although causality is unproven, early glycolytic activation appears linked to regenerative repair. Glycolysis thus emerges as both a biomarker of epithelial fate and a potential therapeutic target to limit CKD progression.