The present study was aimed to examine the role of endothelium in the relaxant effect ofhawthorn fruit extract of Crataegus pinnatifida in four different types of rat arteries, posteriorcerebral communicating artery, right descending coronary artery, common carotid artery, andaorta. In 9,11-dideoxy-11a,9a-epoxy-methanoprostaglandin F2a (U46619)-preconstricted arterialrings except for aorta, the extract produced endothelium-independent relaxations with similarpotency. This relaxation was unaffected by pretreatment with 100 µM NG-nitro-L-arginine methylester (L-NAME, the nitric oxide synthase inhibitor), 3 µM 1H-[1,2,4]oxadiazolo[4,2-a]quinoxalin-1-one (ODQ, the guanylate cyclase inhibitor), or 10 µM indomethacin (the cyclooxygenaseinhibitor). Putative K+ channel blockers (charybdotoxin plus apamin or glibenclamide) did notaffect the extract-induced relaxation in cerebral or coronary artery rings. In contrast, in rat aorticrings the extract produced significantly smaller relaxant response in endothelium-denuded ringsthan that in endothelium-intact rings. Pretreatment with L-NAME or ODQ abolished the extract-induced endothelium-dependent aortic relaxation, whilst indomethacin (3 µM) had no effect. Thepresent results indicate that hawthorn fruit extract possesses a vasorelaxing effect in cerebral,coronary and carotid arteries and this effect is independent of the presence of a functionalendothelium. However, the extract-induced endothelium-dependent relaxation in rat aorta wasmediated through endothelial nitric oxide and cyclic GMP-dependent mechanisms, suggestingthat active components in the extract may act on endothelium to stimulate release of nitric oxidein large conduit arteries of the rats.