In Japan, renal function is estimated based on the serum creatinine (SCr) value, and specifically estimated creatinine clearance (eCCr) obtained with the Cockcroft–Gault (CG) equation and the estimated glomerular filtration rate (eGFRcr) for the Japanese population. However, renal function may be overestimated due to the decrease in SCr value associated with decreased muscle mass. To prevent this, SCr values are rounded up to 0.6 mg/dL for patients when less than 0.6 mg/dL. However, at present, there are few reports providing evidence that predictability has been improved by rounding up SCr values, and thus the scientific basis for rounding up is poor. In this study, we evaluated the validity of SCr values rounded up to 0.6 mg/dL. A retrospective analysis was performed on 11 patients aged 65 years or older with SCr values less than 0.6 mg/dL who were admitted to the Tamana Regional Medical Center in May-August 2017. Measured creatinine clearance (mCCr) was evaluated in 24-h urine samples, and modified glomerular filtration rate (mGFR) was calculated by multiplying mCCr by 0.715. Then, eCCr and eGFRcr using the SCr value measured by the enzymatic method, and eCCr (rounded-up) and eGFRcr (rounded-up) using the SCr value rounded up to 0.6 mg/dL, were compared with mCCr and modified GFR respectively. Correlation and regression analysis, Bland-Altman analysis, and the error index analysis revealed that eCCr (rounded-up) value tended to underestimate renal function more than eCCr value. The eGFRcr value markedly overestimated renal function, and the eGFRcr (rounded-up) value was also overestimated, although to a lesser extent. These results indicate the non-validity of rounding up SCr values to 0.6 mg/dL in both the eCCr and eGFRcr estimation formulas. Thus, unnecessarily rounding up the SCr values should be avoided in sarcopenia patients.