Purpose: In children, 24-hour urine collections are unreliable for evaluating glomerular filtration rate (GFR) because of the difficulty of regulating voiding and the daily variation of urinary creatinine up to 25%. Additionally, creatinine clearance (Ccr) based on urinary creatinine is considered inaccurate. The purpose of this study was to compare estimated GFR determined using Ccr, formulas with serum cystatin C and creatinine, and $^{99m}Tc$-mercaptoacetyltriglycine (MAG3) dynamic renal scintigraphy. Methods: This retrospective study included 101 patients (age, <18 years) who visited Chung-Ang University Hospital between July 2011 and August 2012. GFR was estimated using 24-hour urinary creatinine, five formulas with serum creatinine and cystatin C, and $^{99m}Tc$-MAG3 renal scan. Results: Of the 101 patients, glomerular renal diseases were present in 60 patients (59.4%) and non-glomerular diseases were present in 41 patients (40.6%). There was a significant correlation between estimated GFR determined using $^{99m}Tc$-MAG3 renal scan and Ccr (r=0.389, P <0.001). The correlation values between estimated GFR determined using $^{99m}Tc$-MAG3 renal scan and each formula of Schwartz, Counahan-Barratt, Cockcroft-Gault, Filler and Lepage, and Bokencamp were 0.265 (P=0.007), 0.128 (P=0.044), 0.230 (P=0.021), 0.356 (P<0.001), and 0.355 (P <0.001), respectively. $^{99m}Tc$-MAG3 renal scan was correlated with estimated-GFR by all formulas in decreased renal function. Conclusion: Estimated GFRs determined using serum creatinine and cystatin C, and $^{99m}Tc$-MAG3 renal scan correlated well with estimated GFR determined using Ccr. $^{99m}Tc$-MAG3 renal scan may be replaced for evaluation of renal function with convenience in patients with renal disease and decreased renal function in childhood.