Halenia elliptica D. Don belongs to Gentianaceae family. It is often used as part of a traditional Tibetan medicine to treat hepatitis. In the present investigation, six major xanthone components were isolated and identified from Halenia elliptica. An HPLC/DAD/APCI/MS method was developed and validated for the quantitative analysis of these xanthones, including 1-hydroxy-2,3,5-trimethoxy-xanthone (HM-1), 1-hydroxy-2,3,4,7-tetramethoxy- xanthone (HM-2), 1-hydroxy-2,3,4,5-tetramethoxy-xanthone (HM-3), 1,7- dihydroxy-2,3,4,5-tetramethoxy-xanthone (HM-4), 1,5-dihydroxy-2,3-dimethoxy-xanthone (HM-5) and 1-0-[beta-D-xylopyranosyl-(1-6)-beta-D-glucopyranosyl]-2,3,5-trimethoxy-xanthone (HM-2-10). All the xanthones aglycons caused vasodilation in the coronary artery pre-contracted with 1 muM 5-HT, but the xanthone glycoside had no effect. HM-1 was one of the most abundant xanthones with the most potent vasorelaxant activity. Mechanisms of the vasorelaxant effect of HM-1 were investigated. HM-1 showed a potent vasorelaxant activity on rat coronary artery involved both an endothelium-dependent mechanism involving NO and an endothelium-independent mechanism by inhibiting Ca2+ influx through L-type voltage-operated Ca2+ channels. Taken together, in spite of the pharmacokinetics results showed that HM-1 was rapidly and widely distributed to tissues after intravenous administration in rats, with conjugation to being the major metabolic pathway in vivo, both HM-1 and its active metabolite (HM-5) show that they are important pharmacological agents with potentially useful therapeutic indications. The metabolism and pharmacokinetics of HM-1 displayed biphasic elimination kinetics, with an elimination half-life of 60.4 +/- 4.2 min. Four other Phase I metabolites were isolated and identified as demethylated products in vitro. HM-1 was metabolised to HM-5 in the liver. Biliary excretion studies showed that both HM-1 and the metabolite (HM-5) underwent extensive phase II conjugation to form glucuronides and sulfates. Tissue distribution studies showed that HM-1 was widely distributed to different organs. Collection of urine and faeces over 24 h showed that 10.88% of dose was excreted from urine and 1.91% of dose via faeces. With HM-5 being one of the major in vivo metabolites of HM-1, the effect of HM-5 has been studied on rat coronary artery and compared to HM-1. HM-5-mediated vasorelaxant effect was mediated through opening of potassium channel (TEA, 4-AP) and altering intracellular calcium by partial inhibition of Ca2+ influx through L-type voltage-operated Ca 2+ channels and intracellular Ca2+ stores.