We compared the performance of technetium-99m-diethylenetriaminepentaacetic acid (Tc-99m-DTPA) renal dynamic imaging (RDI), the MDRD equation, and the CKD EPI equation to estimate glomerular filtration rate (GFR). A total of 551 subjects, including CKD patients and healthy individuals, were enrolled in this study. Dual plasma sample clearance method of Tc-99m-DTPA was used as the true value for GFR (tGFR). RDI and the MDRD and CKD EPI equations for estimating GFR were compared and evaluated. Data indicate that RDI and the MDRD equation underestimated GFR and CKD EPI overestimated GFR. RDI was associated with significantly higher bias than the MDRD and CKD EPI equations. The regression coefficient, diagnostic precision, and consistency of RDI were significantly lower than either equation. RDI and the MDRD equation underestimated GFR to a greater degree in subjects with tGFR >= 90 ml/min/1.73 m(2) compared with the results obtained from all subjects. In the tGFR60-89 ml/min/1.73 m(2) group, the precision of RDI was significantly lower than that of both equations. In the tGFR30-59 ml/min/1.73 m(2) group, RDI had the least bias, the most precision, and significantly higher accuracy compared with either equation. In tGFR < 30 ml/min/1.73 m(2), the three methods had similar performance and were not significantly different. RDI significantly underestimates GFR and performs no better than MDRD and CKD EPI equations for GFR estimation; thus, it should not be recommended as a reference standard against which other GFR measurement methods are assessed. However, RDI better estimates GFR than either equation for individuals in the tGFR30-59 ml/min/1.73 m(2) group and thus may be helpful to distinguish stage 3a and 3b CKD.