Because of the high prevalence of chronic kidney disease, estimation of the glomerular filtration rate (GFR) is necessary to diagnose, stage, and follow the progression of renal impairment, and to adjust the dosage of medications with predominantly renal excretion. The main pitfall of using 24-h urinary creatinine clearance is the inaccuracy of urine collection. Multiple formulas based on serum creatinine have been proposed for the estimation of renal function in daily clinical practice and in large-scale studies. The two most widely used formulas are Cockcroft-Gault (CG) for the estimation of creatinine clearance and MDRD (Modification of Diet in Renal Disease) for the estimation of GFR. However, the performance of these formulas is satisfactory only in individuals with a GFR level less than 60 mL/min/1.73 m(2), and the presence of determinants of serum creatinine that are not dependent on GFR, such as gender, age, body weight, or chronic illness, should also be considered. Because of the need for an accurate and reproducible measurement of serum creatinine, uniform creatinine assay calibration is now available. The utility in daily practice of new markers of GFR, such as cystatin C, remains to be demonstrated.