Background: Contrast-induced acute kidney injury (CI-AKI) is a frequent complication after percutaneous coronary intervention (PCI) and severely affects morbidity and mortality. Multiple prediction models for the development of contrast-induced nephropathy (CIN) have been published using heterogeneous characteristics of study populations. Objectives: We sought to compare two contrast-induced acute kidney injury-risk prediction models in patients with reduced glomerular filtration rates undergoing percutaneous coronary interventions. Methods: A cross-sectional study. Results: We evaluated 135 patients who underwent percutaneous coronary intervention from January to May 2017 at Cho Ray Hospital, Ho Chi Minh City, Vietnam. The mean age of the study participants was 68.9±9.9 years and 71.9% were male. The mean baseline creatinine was 1.35±0.35 mg/dl. Contrast-induced acute kidney injury occurred in 18 patients (13.3%). Univariate regression analysis showed that a history of myocardial infarction (OR 0.34), left ventricular ejection fraction (OR 0.94), left ventricular ejection fraction <40% (OR 3.5), hemoglobin (OR 0.98), and Mehran scores (OR 1.16), as well as ACEF-MDRD scores (OR 1.8), were the independent predictors of contrast-induced acute kidney injury. Multivariate regression analysis showed that no single factor was able to predict the incidence of contrast-induced acute kidney injury. Area under curves of Mehran risk scores and ACEF-MDRD risk scores in predicting the incidence of contrast induced acute kidney injury were 0.64 and 0.68, respectively (p=0.61). Conclusions: The study showed that the ACEF-MDRD risk model and Mehran risk model were of similar value in predicting the incidence of contrast-induced acute kidney injury in patients with reduced glomerular rate undergoing percutaneous coronary intervention.