Eight healthy male volunteers received 16 doses of sublingual nitroglycerin tablets (0.4 mg). After 8 minutes, each subject rinsed out his mouth to halt the drug absorption process. The mouth rinses were assayed by high-performance liquid chromatography for residual nitroglycerin content. Each subject also received intravenous infusions of nitroglycerin so that the absolute bioavailability could be evaluated. Plasma nitroglycerin concentrations were determined using a specific and sensitive capillary gas chromatographic method capable of quantifying 25 pg/ml of nitroglycerin. The mean bioavailability (+/- standard deviation) of sublingual nitroglycerin, estimated from plasma concentrations, was 36.2 +/- 24.9% (range 2.6 to 113%). The amount of drug not absorbed after 8 minutes, as determined from the analysis of the mouth rinses, varied from 2.7 to 65.8% (mean 31.4 +/- 18.9%) of the administered sublingual dose. Mean nitroglycerin peak concentrations of 1.89 +/- 1.64 ng/ml were obtained at a mean peak time of 5.3 +/- 2.3 minutes. Thus, sublingual absorption is not instantaneous and can be relatively slow, with peak times of as long as 10 minutes. These data indicate that nitroglycerin pharmacokinetic values should not be estimated only from sublingual doses. Additionally, attempts to correlate pharmacodynamic measurements to sublingual doses must take into account the low and variable bioavailability and the potentially long peak times after sublingual nitroglycerin administration to patients.