In this prospective randomised study 12 patients suffering from cirrhosis of the liver (stable phase) and 12 healthy male volunteers were treated with either 0.3 mg beta-methyldigoxin (Lanitop) or 0.4 mg beta-acetyldigoxin (Novodigal) daily, orally. Every day the total serum digoxin concentrations of the patients and volunteers were measured by radioimmunoassay. Both digoxin and beta-methyldigoxin are measured by this method. In subjects receiving beta-methyldigoxin therapy the ratio of beta-methyldigoxin to digoxin in the serum was determined by liquid chromatography. The digoxin levels in patients with cirrhosis treated with beta-methyldigoxin were statistically significantly higher than in healthy volunteers. In patients with cirrhosis the proportion of serum beta-methyldigoxin averaged 77.7% of the total digoxin concentration, whereas the proportion was only 37.5% in healthy volunteers. With beta-acetyldigoxin there was no statistically significant difference between patients with cirrhosis and healthy volunteers. Alterations in pharmacokinetics may cause the higher total serum digoxin concentrations in cirrhotic patients. The following factors seem to be important: longer elimination half life, changes in distribution volume and reduced renal clearance. There is greater danger of digitalis toxicity in patients with cirrhosis of the liver on standard dosage of beta-methyldigoxin than on standard dosage of beta-acetyldigoxin.