The influence of clofibrate on cholesterol metabolism in patients with hyperlipidemia was studied by means of sterol balance and isotope kinetic techniques and by measurements of flow rates of cholesterol through the biliary tract. Long-term balance studies were carried out on a metabolic ward in 24 patients with all currently recognized types of hyperlipidemia; in five other patients with hypercholesterolemia, pool sizes and turnover rates of cholesterol were defined by compartmental analysis before and after three years' daily administration of the drug. Except in fat-induced hypertriglyceridemia (two patients), clofibrate caused reduced plasma levels of triglycerides and cholesterol in all categories of hyperlipidemia. As a general rule, excretion of cholesterol into bile and feces was significantly increased and fecal bile acid excretion was decreased, regardless of the type of lipoprotein abnormality. Despite a net increase in steroid excretion in most patients with hyperlipidemia, cholesterol synthesis was not increased; indeed, in many patients synthesis appeared to be decreased. While the data obtained in 29 patients were not always consistent, the bulk of the evidence suggests that, in all forms of hyperlipidemia except fat-induced hyperglyceridemia, the drug causes an increased output of cholesterol while simultaneously inhibiting any compensatory increase in cholesterol synthesis. Therefore, it appeared that the increased excretion of steroids was most likely derived from cholesterol stored in tissues. This conclusion was strengthened by finding that long-term administration of the drug can cause marked reduction in body pools of cholesterol. These findings are reflected clinically by resolution of skin and tendon xanthomatosis. However, it is not yet known whether the accumulation of cholesterol in arterial walls that is part of the process of atherogenesis can be inhibited or reversed by the drug.