This study compares pharmacokinetic parameters and colonic tissue concentrations of cyclosporine administered by olive-oil or water-retention enemas with conventional intravenous (i.v.) and oral dosing. Five medical students were enrolled in a prospective crossover study. All subjects received a single dose of cyclosporine on four separate occasions, once orally, once as an olive-oil enema, once as a water enema, and once i.v. Cyclosporine concentration was measured in blood and in colonic tissue obtained by flexible sigmoidoscopy. Bioavailability was 18 +/- 7% (mean +/- SD) for the oral dose and was unmeasurable for the oil and water enemas. The concentration of cyclosporine in colon tissue was 32,443 +/- 17,251 ng/g (mean +/- SD) for the i.v. dose, 2797 +/- 1812 ng/g for the oral dose, 21,727 +/- 14,090 ng/g for the oil enema, and 25,318 +/- 30,408 ng/g for the water enema. The authors conclude that the bioavailability of cyclosporine, and thus the systemic absorption after administration by a retention enema, is negligible. The colonic tissue concentration of cyclosporine after i.v. or rectal administration via an enema is tenfold higher than that for oral dosing. These findings suggest that cyclosporine-retention enemas produce high distal colonic tissue concentrations with negligible systemic absorption after a single dose in healthy subjects and should be evaluated as treatment for patients with left-sided colitis. Because cyclosporine administered by the i.v. route provided sharply higher colonic tissue concentrations than those seen with oral therapy, pulse i.v. cyclosporine should be tried for patients with severe ileitis and colitis.