Here we compare the Reference Method (I) and the Definitive Method (II) for measurement of cholesterol in serum. For pure cholesterol solutions, values by I agree with values by II, but for fresh, frozen, or lyophilized sera, values by I average 1.6% higher than values by II. We conclude that an undetected interference is associated with I. However, the observed difference does not diminish the usefulness of I as an accuracy base for cholesterol measurements, because it is clinically insignificant at present.