【Abstract】Background The RIFLE classification scheme for the evaluation of acute kidney injury (AKI) is based on relative changes in serum creatinine (SCr) and on urine output. The SCr criteria, therefore, require a pre-morbid baseline value. When the baseline SCr is unknown, it is currently estimated by using the MDRD equation. However, the MDRD approach assumes a glomerular filtration rate of 75 mL/min/1.73 m2, which has been seldom validated. Methods Patients undergoing coronary artery bypass graft from January 2007 to January 2009 were retrospectively evaluated. The RIFLE class was determined by using the observed pre-operation SCr values and the estimated baseline SCr values. The diagnosis sensitivity and specificity for prediction of mortality in hospital were compared between the RIFLE criteria based on the observed pre-operation values and the estimated baseline SCr values. Results One hundred and nineteen patients were included in this study. There was no significant difference in the prevalence of AKI between patients after coronary artery bypass graft diagnosed by RIFLE criteria based on the observed pre-operation values and those based on the estimated baseline SCr values (44.50% vs. 43.70%, P > 0.05). Logistic regression analysis suggested that the relative risk of hospital mortality for AKI was similar between patients diagnosed by RIFLE criteria based on the observed pre-operation values and those based on the estimated baseline SCr values. The area under the ROC curve for hospital mortality was 0.829 for RIFLE based on the observed pre-operation SCr values and 0.853 for RIFLE based on the estimated baseline SCr values (P>0.05). Conclusions When pre-operation SCr value is unavailable, RIFLE criteria based on the estimated baseline SCr by the MDRD equation can be used for the AKI diagnosis in patients after coronary artery bypass graft with nearly normal pre-operation glomerular filtration rate.