Liver haemodynamics and liver function were measured in 34 alcoholic cirrhotic men before entry and after 12 months (median) in a double-blind, placebo-controlled study on the effect of oral testosterone treatment (200 mg t.i.d.). Comparing data at entry with those at follow-up in the total patient group, a significant change in median values of portal pressure (-23%, n = 34, P less than 0.005), hepatic blood flow (-22%, n = 28, P less than 0.001), indocyanine green clearance (+16%, n = 29, P less than 0.01), and galactose elimination capacity (+8%, n = 31, P less than 0.05) was observed. However, testosterone-treated patients did not differ significantly from placebo-treated patients regarding any of the measured variables. No significant relationships could be demonstrated between ethanol consumption and liver haemodynamics and liver function, but the number of patients consuming more than 100 g ethanol per day decreased significantly (P less than 0.001) from 22 (65%) before entry to one (3%) during follow-up. In conclusion, oral testosterone treatment of men with alcoholic cirrhosis does not explain the significant improvement of liver haemodynamics and function observed in this study. However, the improvement may be due to reduced ethanol consumption.