Lysyl oxidase was partially purified from serum by a diethylaminoethyl batch procedure in the presence of 6 mol/L urea and dialyzed against 3 mol/L KSCN. Using this method, we determined serum lysyl oxidase activity in 52 patients with liver disease and in 14 healthy controls, and we examined usefulness of serum lysyl oxidase in assessing liver fibrogenesis. For this purpose, serum lysyl oxidase activity in chronic liver disease was compared with serum levels of prolyl hydroxylase and laminin P1. As compared with controls, serum lysyl oxidase activity increased 1.6-fold in chronic persistent hepatitis, 4.4-fold in chronic active hepatitis and 11.8-fold in cirrhosis, indicating an increase in concert with the development of liver fibrosis. In hepatocellular carcinoma, the serum activity, although significantly increased, was lower than that in cirrhosis. Serum prolyl hydroxylase was significantly increased in chronic active hepatitis, in liver cirrhosis and in hepatocellular carcinoma. Serum laminin P1 was significantly increased in chronic active hepatitis, in cirrhosis and in hepatocellular carcinoma. Serum lysyl oxidase activity did not correlate significantly with serum levels of prolyl hydroxylase and laminin P1 in any subject or in any subgroup. The magnitude of the increase and the abnormal percentage of serum lysyl oxidase activity were larger than those for serum prolyl hydroxylase and laminin P1. These results suggest that serum lysyl oxidase activity is a more sensitive indicator of liver fibrosis than serum prolyl hydroxylase and laminin P1.