The pharmacokinetic behavior of caffeine was compared in a group of eight healthy young men aged 20.5 +/- 2.0 years (mean +/- SD), and in a group of eight healthy, elderly men aged 71.2 +/- 3.9 years. Each subject was given a 5 mg/kg dose of caffeine as either an aqueous oral solution or an intravenous infusion over 30 min using a randomized crossover design. Plasma and urine samples were collected for 24 hr following each dose and analyzed for caffeine content using high-performance liquid chromatography. The peak times (tmax), peak concentrations (Cmax), and the percentage of the peroral dose systemically available, F(%), were essentially identical in both age groups, indicating that caffeine was absorbed rapidly and completely after peroral administration. These results also indicated that the first-pass metabolism observed in rats following the peroral administration of caffeine does not occur in either human group studied here. The elimination of caffeine during its terminal disposition phase was log-linear. Several between-group comparisons of other pharmacokinetic parameters were made. Although the average elimination rate constant was greater in the elderly, the difference did not reach statistical significance, possibly because of the considerable intersubject variability in the elimination rate of caffeine, with half-lives ranging from 2.27 to 9.87 hr. The average apparent volume of distribution was significantly lower in the elderly subjects while the clearances were slightly, but not significantly, larger in the elderly subjects. It appears that most aspects of the pharmacokinetic behavior of caffeine are very similar in young and elderly men.