We measured des-gamma-carboxyglutamic acid prothrombin (protein induced by vitamin K absence or antagonist-Factor II: [PIVKA-II]) in plasmas of normal subjects, patients with thrombotic disease, those with hepatic disease including hepatocellular carcinoma, and those with carcinoma of other tissues, and compared the results with results of blood coagulation tests used for the examination of hepatic function. In addition, in the patients with hepatic disease, PIVKA-II and alpha-fetoprotein (AFP) levels were compared. The PIVKA-II level was frequently high in patients with thrombotic disease given warfarin therapy and those with hepatocellular carcinoma. However, in patients with thrombotic disease who were not given warfarin therapy, no significant correlation was seen between the PIVKA-II value and the results of the thrombotest or hepaplastin test, suggesting no association between the PIVKA-II level and the degree of impairment of hepatic function. In 70 patients with hepatocellular carcinoma, the percentage of patients positive for PIVKA-II (greater than or equal to 0.1 micrograms/ml) and those positive for AFP (greater than or equal to 20 ng/ml) were similar (77% and 74%, respectively). Pearson's correlation of coefficient between the PIVKA-II value and the AFP value in the 70 patients was 0.463. However, false-positive rates in patients with hepatic disease other than hepatocellular carcinoma were lower for PIVKA-II. Combined assessment of PIVKA-II and AFP increased positive rates and allowed exclusion of false-positive patients. The plasma PIVKA-II level is suggested to be useful as an indicator of warfarin control in patients with thrombotic disease, as a marker of hepatocellular carcinoma, and is particularly of value when assessed in combination with AFP.