OBJECTIVE To evaluate the pharmacokinetics and pharmacodynamics of furosemide and torasemide in patients with cirrhosis and diuretic resistant ascites. METHODS Eighteen patients were randomly allocated to receive intravenous torasemide (40 mg) or furosemide (80 mg). The renal response to these drugs was assessed in baseline conditions and in the 24 h following drug administration together with plasma and urinary concentrations of furosemide, torasemide and its metabolites. RESULTS Torasemide induced significantly greater diuretic and natriuretic effects than furosemide in the first hour after drug administration. No other significant differences between the two drugs were observed with respect to the renal response to these drugs. Torasemide reached a lower maximum plasma concentration than furosemide, but the former drug had a longer apparent terminal half-life and lower renal and non-renal clearances. Comparing these results with those previously reported in healthy subjects, both drugs showed a reduced elimination rate through renal and non-renal routes, and a larger distribution to body fluids. As a consequence, the half-life of both drugs was longer than in healthy subjects. Urinary excretion of pharmacologically active species, however, was quantitatively unchanged after torasemide administration, whereas it was reduced after furosemide. Finally, the natriuretic potency of both drugs was markedly reduced in these patients. CONCLUSIONS The pharmacokinetics and pharmacodynamics of torasemide and furosemide are markedly altered in patients with diuretic resistant ascites.