The plasma level of low-density lipoprotein (LDL) cholesterol is the "gold standard" for estimating the lipoprotein-related risk for complications of atherosclerotic vascular disease. LDL cholesterol concentrations are commonly estimated by the Friedewald formula that requires only the measurement (after overnight fasting) of plasma cholesterol and triglycerides along with high-density lipoprotein (HDL) cholesterol. This value, however, is not in fact a true estimate of LDL cholesterol but rather of LDL cholesterol along with variable, usually smaller, amounts of intermediate-density lipoprotein (IDL) cholesterol and lipoprotein(a). Estimation of LDL cholesterol levels by the Friedewald formula becomes progressively less accurate as plasma triglyceride concentrations increase, and the formula is generally considered inapplicable when triglyceride levels exceed 400 mg/dL. We believe that a very simple measurement-non-HDL cholesterol (serum cholesterol minus HDL cholesterol)-has considerable potential as a screening tool for identifying dyslipoproteinemias, for risk assessment, and for assessing the results of hypolipidemic therapy. Unlike the estimation of LDL cholesterol levels by the Friedewald formula, the estimation of non-HDL cholesterol concentrations requires no assumptions about the relation of very-low-density (VLDL) cholesterol levels to plasma triglyceride concentrations. This method includes all of the cholesterol present in lipoprotein particles now considered to be potentially atherogenic [VLDL, IDL, LDL, and lipoprotein(a)]. This article provides examples of the utility of non-HDL cholesterol concentrations in clinical medicine.