The present report has presented the first clear evidence in man for the existence of specific hepatic cholesterol precursor sites associated with the formation and secretion of bile acids and biliary cholesterol. These hepatic compartments derive virtually all their cholesterol from newly synthesized and lipoprotein free cholesterol. The model which is presented was formulated on current concepts of cholesterol metabolism in man and is concerned, at this initial stage, with the elucidation of the bile acid and biliary cholesterol compartments. The complexity of cholesterol metabolism in man necessitated an initial approach that would minimize the number of inputs of cholesterol into the system, allow for the sampling of several cholesterol compartments, and permit the simultaneous labeling of newly synthesized cholesterol and preformed cholesterol. To achieve these objectives, we studied the patient with a total bile fistula. Six patients were administered simultaneously pulse injections of labeled mevalonic acid and [(14)C]cholesterol. The qualitative features of the specific activity time course curves after labeled mevalonic acid revealed no precursor-product relationship between bile acid, biliary cholesterol, and plasma free cholesterol. The peak specific activity of the bile acids was reached in approximately 100 min and was higher than the biliary cholesterol, which was higher than the plasma free cholesterol. The plasma free cholesterol specific activity became higher than the other lipids after 12 h and remained higher throughout the period of study. Similar related observations were made with [(14)C]cholesterol. The data were then subjected to simulation analysis and modeling using the SAAM-27 computer program. Computer least-square fits of the data were obtained after the model was evolved. During the model development, the least number of compartments and transport pathways were introduced consistent with a good fit of the data. Of particular importance was the constraint that the model fit the data obtained from both [(14)C]cholesterol and labeled mevalonic acid. The same parameter values were used to fit the data from both tracers. The fluxes arrived at in the model indicate that 31% and 20%, respectively, of the cholesterol input into the bile acid and biliary cholesterol precursor sites were derived directly from the newly synthesized hepatic cholesterol. The remainder had its origin predominantly from lipoprotein free cholesterol. Plasma esterified cholesterol (as free) made a small contribution (11%) to the bile acid compartment. Similarly, 10% of the biliary cholesterol arose from an unknown hepatic site. The present report has provided the basis for a new procedure for studying in vivo cholesterol metabolism in man. Examination of the derived cholesterol flux rates between the compartments suggests the presence of an important mechanism regulating the partitioning of lipoprotein free cholesterol between the bile acid and biliary cholesterol precursor sites. Aberrations in the proportioning of precursor cholesterol between these sites could be a causative factor precipitating the excessive secretion of biliary cholesterol and the production of lithogenic bile.