A randomized placebo-controlled 2-year study was performed in 69 patients with primary biliary cirrhosis (PBC) on serum lipids during ursodeoxycholic acid (URSO) and colchicine treatments. In addition to serum bilirubin and alkaline phosphatase (AFOS), two variables considered to reflect liver function, serum lipoproteins, cholesterol precursors (squalene, Δ8-cholestenol, lathosterol and desmosterol), markers of cholesterol synthesis, cholestanol and plant sterols (campesterol and sitosterol), markers of liver function and cholesterol absorption, were studied before and during the treatments. Serum bilirubin was inconsistently improved by URSO, whereas improvement of AFOS values was better by URSO than colchicine, especially in patients with initially more advanced PBC. Serum total cholesterol was reduced by both drugs, very low—density lipoprotein (VLDL) and high-density lipoprotein (HDL) cholesterol by URSO. Cholesterol precursor sterols were increased by both URSO and colchicine mainly in patients with initially less severe PBC. On the other hand, the cholestanol values were markedly increased initially, and the values were related to bilirubin during the 2-year period, were further increased in the placebo group, and reduced in the URSO and colchicine groups, so that the improvement was highest in the URSO-treated patients with the severe form of PBC. The increase of the serum plant sterols, particularly that of sitosterol, was retarded by the two drugs so that the campesterol/sitosterol ratio, which was related to serum bilirubin, was increased especially in the cases with initially more advanced PBC. The present findings show that the serum cholestanol level and the campesterol/sitosterol ratio, two variables characterizing end-stage PBC, are modestly but significantly improved by long-term URSO or colchicine treatment, but virtually none of the excessively increased cholestanol values are normalized.