Abstract: Objective：To explore the relationship between elevated perioperative red blood cell distribution width (RDW) and cardiac surgery-associated acute kidney injury (CSA-AKI). Methods：Clinical data of patients underwent cardiac surgery during January 2010 to December 2014 at the Cardiac Surgery Department of Zhongshan Hospital, Fudan University were retrospectively analyzed. The elevated RDW was defined as the difference between RDW 24 hours after cardiac surgery and the latest RDW before cardiac surgery. Bivariate correlation analysis were used to screen variates associated with the elevated RDW. Propensity score matching (1∶1) was used on the basis of clinical characteristics and preoperative variables, in order to decrease the confounding bias. Logistic regression were performed in the matched cohort to screen the predictors for CSA-AKI. Results：A total of 9 453 eligible patients were enrolled and 5 690 patients were matched. In the entire cohort, the AKI incidence was 34.32% (3 244/9 453), in-hospital mortality was 2.25% (213/9 453), renal replacement therapy (RRT) rate was 1.64% (155/9 453). The elevated value of RDW in patients who developed AKI was significantly higher than that in patients who didn’t (0.67±1.09 vs 0.37±0.88, P<0.01). Logistic regression analysis showed that aging (OR=1.035), increased pre-operative serum creatinine (OR=1.012), male (OR=1.528), prolonged cardiopulmonary bypass elapse (OR=1.009), hypertension (OR=1.236), diabetes (OR=1.290), pre-operative NYHA 3-4 (OR=1.199), valve combined coronary bypass surgery (OR=2.559), valve surgery (OR=1.745), aortic dissection surgery (OR=4.754), and elevated RDW (OR=1.315) were predictors of CSA-AKI. Conclusions：As a laboratory index, the elevated RDW is convenient, objective, and could predict CSA-AKI, therefore might be beneficial for the early diagnosis, treatment, and prognosis improvement.