Epidemiological and experimental studies suggest that consumption of garlic may protect against several types of cancer. Moreover, a plausible hypothesis has been proposed that the biological effects of garlic can be attributed to the enhancing action of a variety of organosulfur compounds, present in garlic, on hepatic phase II carcinogen detoxification enzymes. We have used the N-methylnitrosourea (NMU)-induced rat mammary tumor model to test the chemopreventive effects of a water-soluble organosulfur constituent derived from aged garlic, S-allylcysteine (SAC). Rats were fed diets supplemented with 666 and 2,000 ppm SAC beginning seven days before initiation with NMU (55 days of age) to termination (18 wk post-NMU), at which time mammary tumors were enumerated. At neither dose did SAC exert an inhibitory effect on any index of tumor development, including incidence, latency, multiplicity, or volume, compared with untreated controls. Weight gains in all groups were similar. Assay of serum SAC levels in supplemented groups indicated that SAC concentrations were beneath the limits of detection of the high-performance liquid chromatography system used. These results contradict previous animal model studies indicating that SAC acts as an inhibitory agent in experimental mammary tumorigenesis; reasons for this discrepancy include the possibility that SAC may exhibit nonlinear dose effects.