The American Academy of Pediatrics' Committee on Nutrition believes that a child with a family history of either early coronary heart disease or hyperlipidemia should have a complete evaluation, including serum or plasma measurement of cholesterol, other lipids, and lipoproteins. On the other hand, the Committee on Nutrition does not favor universal testing of blood cholesterol levels of children either in the hospital or in the office of the private physician. At this time, the value of routine testing appears outweighed by the disadvantages; office testing has additional limitations. We believe that universal cholesterol testing should not be part of continuing well-child care for the following reasons. 1. The standardization of equipment and the measurement of cholesterol are fraught with difficulties. Precision and accuracy are essential cornponents of any laboratory measurement. Survey data show that replicate measurements of a sample containing 240 mg/dL of cholesterol resulted in values ranging from 192 to 288 mg/dL. Because a low-fat diet could only be expected to decrease the cholesterol level 10% to 15% (204 to 216 mg/dL), an improvement could thus be obscured by the error of the measurement. 2. Many of the instruments suggested for the measurement of cholesterol in the office have not been adequately standardized for field conditions. Cholesterol measurements obtained from an assortment of instruments available commercially ranged from 187 to 379 mg/dL (the reference standard was 262 mg/dL). Hospital laboratories are often precise but not always accurate; office determinations frequently lack both precision and accuracy. Moreover, proper performance of this laboratory procedure requires technical skill not always available in the office.