1, 5-Dihydroxy-2, 3-dimethoxy-xanthone (HM-5) is one of the naturally-occurring xanthones of a Tibetan medicinal herb Halenia elliptica. Recently, it has been shown that HM-5 is one of the phase I metabolites of 1-hydroxy-2, 3, 5-trimethoxy-xanthone (HM-1), the major active component of H. elliptica with potent vasorelaxant actions. This study investigated the vasorelaxant effect of HM-5 and its mechanism(s). HM-5 (0.35–21.9 μM) produced a concentration-dependent relaxation in rat coronary artery rings pre-contracted with 1 μM 5-hydroxytryptamine (5-HT), with an EC 50 of 4.40 ± 1.08 μM. Unlike HM-1, the effect of HM-5 was endothelial-independent such that removal of the endothelium did not affect its vasodilator potency. Nitric oxide synthase (NOS) inhibitor N ω-nitro- l-arginine methyl ester ( l-NAME, 100 μM), the soluble guanylate cyclase inhibitor 1 H-[1,2,4] oxadiazolo [4,3-α] quinoxalin-1-one (ODQ, 10 μM) did not affect the vasodilatory effects of HM-5, thus confirming the non-involvement of endothelium related mechanisms. In endothelium-denuded coronary artery rings, the vasorelaxant effect of HM-5 was inhibited by a potassium channel blocker, TEA (10 mM), and 4-aminopyridine (4-AP, a K v blocker; 1 mM) but not by other K + channel blockers such as iberiotoxin (100 nM), barium chloride (100 μM) and glibenclamide (10 μM). The involvement of Ca 2+ channel was studied in artery rings pre-incubated with Ca 2+-free buffer (intact endothelium or endothelium-denuded) and primed with 1 μM 5-HT or 60 mM KCl prior to the addition of CaCl 2 to elicit contraction. In the 5-HT-primed preparations, HM-5 (34.7 μM) significantly inhibited the CaCl 2-induced vasoconstriction (89.9% inhibition in intact endothelium artery rings; 83.3% inhibition in endothelium-denuded rings). In the KCl-primed preparations, HM-5 (34.7 μM) produced a 34% inhibition in endothelium-denuded rings. The same concentration of HM-5 inhibited (by 62.3%) the contractile response to 10 μM phorbol 12, 13-diacetate (PDA), a protein kinase C activator, in Ca 2+-free solutions. Taken together, this study showed that the mechanisms of the vasorelaxant effects of HM-5 were distinctly different from those of its parent drug HM-1. The vasorelaxant effect of HM-5 was mediated through opening of potassium channel (4-AP) and altering intracellular calcium by partial inhibition of Ca 2+ influx through L-type voltage-operated Ca 2+ channels and intracellular Ca 2+ stores.