The precise and accurate measurement of total cholesterol is necessary to correctly identify hypercholesterolemia. The aim of this study was to assess the measurement of serum cholesterol in clinical laboratories using as acceptability criteria the recommendations of the ''National Cholesterol Education Program'' (that consider acceptable a deviation less-than-or-equal-to 5% from the real value and ideal a deviation less-than-or-equal-to 3%). In each of three assessments, three pools of lyophilized sera with different levels of cholesterol were used method. Twenty nine to 42.3% of laboratories had results within the ideal interval and 49.5 to 60% within the acceptable range of deviation In the last assessment, only 17% of laboratories had their three values within the ideal range Only 39.4% of laboratories use 200 mg/dl as the cut off point to diagnose hypercholesterolemia. It is concluded that the standardization of cholesterol measurement and the use of a common cut off point to diagnose hypercholesterolemia must be emphasized.