Three groups of 3 rats received oral doses (8 mg/kg) of garlic constituents (alliin, allicin and vinyldithiines (2-vinyl-[4H]-1,3-dithiine and 3-vinyl-[4H]-1,2-dithiine)) in the form of an oil macerate of the 35S-labeled substance. The measured activity was referred to 35S-alliin (35S-alliin equivalents). The blood activity levels in each group were monitored for 72 h. For 35S-allicin and the labeled vinyldithiines the excretion with the urine, feces, and exhaled air was also measured. The distribution among the organs (whole-body autoradiography) and the urinary metabolite pattern (thin-layer chromatography) were also determined. For 35S-alliin the blood activity profile differed considerably from those of 35S-allicin and the labeled vinyldithiines: both the absorption and the elimination of the radioactivity were distinctly faster than for the other garlic constituents, maximum blood levels being reached within the first 10 min and elimination from the blood being almost complete after 6 h. For the other garlic constituents the maximum blood levels were not reached until 30-60 min (35S-allicin) or 120 min (vinyldithiines) p.a. and blood levels > 1000 ng-Eq/ml were still present at the end of the study after 72 h. The mean total urinary and fecal excretion after 72 h was 85.5% (35S-allicin) or 92.3% (labeled vinyldithiines) of the dose. The urinary excretion indicates a minimum absorption rate of 65% (35S-allicin) or 73% (vinyldithiines). It is uncertain whether the 19-21% recovered in the feces was unabsorbed substance or had been excreted via the bile or intestinal mucosa. The exhaled air showed only traces of activity although the whole-body autoradiographs, after fairly long exposure (96 h), showed distinct enrichment of activity in the mucosa of the airways and pharynx. The activity is deposite mainly in the cartilage of the vertebral column and ribs. There was no detectable difference in organ distribution between 35S-allicin and the labeled vinyldithiines. All that could be established from the urinary metabolite pattern was that unchanged 35S-allicin and unchanged labeled vinyldithiines are absent. There is therefore extensive metabolization. The metabolites must have a very polar structure with acid functional groups since satisfactory separation was achievable only with acid solvent systems. Conjugates with sulfuric or glucuronic acid were not detectable. These results reveal no differences in pharmacokinetic behavior between 35S-allicin and the labeled vinyldithiines. A final verdict as to whether the metabolites, which may be pharmacologically active, are identical must await further studies designed to identify the metabolites.