(-)-Epigallocatechin gallate (EGCG), the most abundant tea catechin, has been proposed to be beneficial to human health based on its strong antioxidative and other biological activities in vitro. Inadequate knowledge regarding the bioavailability and biotransformation of EGCG in humans, however, has limited our understanding of its possible beneficial health effects. In this study, 4',4' '-di-O-methyl-EGCG (4',4' '-DiMeEGCG) was detected in human plasma and urine by LC/MS/MS following green tea ingestion. Both 4',4' '-DiMeEGCG and EGCG reached peak plasma values (20.5 +/- 7.7 and 145.4 +/- 31.6 nM, respectively, in 4 subjects) at 2 h after the dose. The half-lives of 4',4' '-DiMeEGCG and EGCG were 4.1 +/- 0.8 and 2.7 +/- 0.9 h, respectively. The cumulative urinary excretion of 4',4' '-DiMeEGCG during a 24 h period was 140.3 +/- 48.6 microg, about 5-fold higher than that of EGCG, but the excreted 4',4' '-DiMeEGCG and EGCG in urine only accounted for about 0.1% of ingested EGCG. (-)-5-(3',4',5'-Trihydroxyphenyl)-gamma-valerolactone (M4) and (-)-5-(3',4'-dihydroxyphenyl)-gamma-valerolactone (M6), along with another possible ring-fission metabolite, (-)-5-(3',5'-dihydroxyphenyl)-gamma-valerolactone (M6'), were detected in human urine after green tea ingestion. The cumulative excretion of M4, M6', and M6 during a 24 h period ranged from 75 microg to 1.2 mg, 0.6 to 6 mg, and 0.6 to 10 mg, respectively. The combined excretion of all three ring-fission metabolites accounted for 1.5-16% of ingested catechins. M4, M6', and M6 were all observed after the ingestion of pure EGCG or EGC by human subjects, whereas only M6 was produced after EC ingestion. These metabolites as well as monomethylated EGCG were detected in mice and rats after tea or EGCG administration, and the tissue levels reflected the rather low bioavailability of EGCG in rats. The presently characterized methylated EGCG metabolites and ring-fission products exist in substantial quantities and may contribute to the biological activities of tea.