Background This study aimed to further evaluate the accuracy of eleven GFR equations in different subgroups of an elderly Chinese hospitalized population. Methods All participants of the study were divided into seven separate groups including age-subgroup, sex-subgroup, GFR Staging-subgroup and whether combined with diabetic, hypertensive, coronary heart disease (CHD) and cerebrovascular disease. Referring to Tc-99m-DTPA dual plasma sample clearance method, six serum creatinine (Cr)-based [Cockcroft-Gault (CG), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPICr), Lund-Malmo Revised (LMR), Berlin Initiative Study (BIS1), Full Age Spectrum (FAS(Cr)) and European Kidney Function Consortium (EKFC)], two serum cystatin C(Cys)-based (CKD-EPICys and FAS(Cys)), and three Cr-Cys combination based (CKD-EPICr-Cys, BIS2 and FAS(Cr-Cys)) equations were employed. Bias, interquartile range of the median difference (IQR), P30, and GFR misclassification rate were calculated to compare the performance of the selected equations. Results A total of 359 elderly Chinese patients were enrolled. Overall, median mGFR was 36.91(25.26,56.32)ml/min/1.73 m(2). Smaller biases (ml/min/1.73 m(2)) were shown in CKD-EPICr and BIS1 equations (0.75 and 0.61). IQR (ml/min/1.73m(2)) was least with BIS2 equation and FAS(Cr-Cys) equation (10.34 and 10.65). For accuracy (P30), performance of FAS(Cr-Cys), BIS2, and BIS1 equation was superior (78.3%, 78.0%, and 74.7%, respectively). In terms of RMSE (ml/min/1.73 m(2)), BIS1 and FAS(Cr-Cys) equation performed better (12.44 and 12.51). Conclusions Overall, this study showed that the eGFR equations were less accurate in the diabetic and non-hypertension group than in the non-diabetic and hypertension group, respectively. Among all enrolled equations, the BIS2 and FAS(Cr-Cys) equations might be the best choice to evaluate glomerular filtration rate in Chinese elderly patients.