Pre-operative evaluation of the quantity of functional remnant hepatic parenchyma after hepatectomy was carried out to predict the optimal amount of hepatic resection using the trimethadione (TMO) tolerance test. This test is an estimate based on serum diamethadione (DMO)/TMO ratio (DMO is the only metabolite of TMO) 4 h after oral administration of TMO, and computed tomography (CT) scans of the liver in patients with hepatocellular carcinoma (HCC). The percentage of remnant hepatic parenchyma was calculated pre-operatively from the remaining non-cancerous portion and the whole hepatic parenchyma, excluding the HCC portion, using the CT scans. Presumptive remnant DMO/TMO ratio was calculated by multiplying the percentage of remnant hepatic parenchyma and pre-operative serum DMO/TMO ratio in 45 patients with HCC undergoing resection. The presumptive remnant DMO/TMO ratios were 0.31 +/- 0.10 (mean +/- s.d.) in 42 patients who survived hepatectomy and 0.13 +/- 0.02 (below 0.15) in the three patients who died from postoperative hepatic failure. Of these surviving patients, two patients who had the presumptive remnant DMO/TMO ratios under 0.15 developed postoperative severe complications. Thus, hepatectomy may not be indicated in patients where presumptive remnant DMO/TMO ratio is 0.15 and lower. These findings suggest that pre-operative measurement of presumptive remnant DMO/TMO ratio, when obtained with the CT of the liver and TMO tolerance test, is useful for prediction of the quantity of functional remnant hepatic parenchyma after hepatectomy in patients with chronic liver disease.