This study investigates the separate effects of age and hepatocellular liver disease on the disposition and elimination of diazepam (Valium) in man. The drug was given either by rapid intravenous injection (0.1 mg/kg) or orally (10 mg) to 33 normal volunteers rnaging in age from 15 to 82 yr as well as to 9 individuals with alcoholic cirrhosis, 8 with acute viral hepatitis, and 4 with chronic active hepatitis. In the normal individuals, the terminal plasma half-life of diazepam, (t 1/2 (B)) exhibited a striking age-dependence; at 20 yr the t 1/2 (beta) was about 20 h, but it increased linearly with age to about 90 h at 80 yr. The plasma clearance of diazepam in the majority of the normal subjects was between 20 and 32 ml/min and showed no significant age-dependence. Cigarette smoking did not affect the half-life or the clearance. Additionally, neither the plasma binding (97.4 plus or minus 1.2%, mean plus or minus SD) nor the blood/plasma concentration ratio (0.58 plus or minus 0.16) of diazepam showed any age-related changes (P greater than 0.05). By contrast, analysis of the intravenous data according to a two-compartment open model indicated that both the initial distribution space (V1) and the volume of distribution at steady state [Vd(ss)] of diazepam increased linearly with age (P less than 0.005). The increase in Vd(ss) was secondary to the change in V1. It appears then that the prolongation of t 1/2 (beta) of diazepam with age is primarily dependent on an increase in the initial distribution volume of the drug. The plasma concentration/time course of the metabolite, desmethyldiazepam, was also affected by age. In older individuals, the initial presence and the peak values of desmethyldiazepam were observed later and the metabolite was present in lower concentrations. Despite the profound prolongation of t 1/2 (theta) with age, the constancy of diazepam clearance indicates that drug plasma concentrations will not accumulate any more in the old than the young, and chronic dosage more in the old than the young, and chronic dosage modifications based on pharmacokinetic considerations are unnecessary. Data obtained in patients with liver disease were compared with those found in age-matched control groups. Patients with cirrhosis showed a more than twofold prolongation in the half-life of diazepam (105.6 plus or minus 15.2 vs. 46.6 plus or minus 14.2 h, P less than 0.001).