Aims The study is aimed at studying the incidence of acute kidney injury (AKI) and exploring the potential predictor for AKI in patients with acute pancreatitis. Methods A retrospective study adopting a stratified cohort sampling design was performed in a cohort of patients (n = 237) diagnosed with acute pancreatitis without any renal injury. The following information including age, gender, serum creatinine, serum urea nitrogen, serum uric acid, serum cystatin C, fasting serum glucose, serum amylase, serum lipase, serum choline esterase, total protein, albumin, globulin, total bilirubin, direct bilirubin, total bile acids, glutamic-pyruvic transaminase, glutamic-oxaloacetic transaminase, gamma glutamyl transpeptidase, and alkaline phosphatase were collected from each patient when they were diagnosed with acute pancreatitis. Student t-test was conducted to figure out the difference between patients with and without AKI. Univariate and multivariate logistic regression analyses were used for investigating the predictors for AKI in patients with acute pancreatitis. Results 18 (7.6%) patients in total had developed AKI among the study group. Compared with patients without AKI (1.01 ± 0.26 mg/L), the level of baseline serum cystatin C (CYS-C) was significantly higher in patients with AKI (3.64 ± 2.17 mg/L, P < 0.001). Baseline serum CYS-C (OR = 203.594, P < 0.001) was the independent and significant predictor for AKI in patients with acute pancreatitis. AKI in patients with acute pancreatitis could be identified with a sensitivity of 88.9% at specificity of 100% (AUC = 0.948, 95% CI 0.879–1.000) by baseline serum CYS-C (cut-off value = 1.865 mg/L). Conclusions Baseline serum CYS-C shall be adopted to predict the potential risk of AKI in patients with acute pancreatitis.