Hypoxic pulmonary vasoconstriction (HPV) is the rapid, reversible increase in pulmonary vascular resistance that occurs when the alveolar oxygen tension falls below a threshold level. HPV demonstrates a transient constriction (phase I) superimposed on a sustained constriction (phase II). An interval of 60 minutes between repeated hypoxic mediums was employed in our protocol to determine the regulation mechanism of HPV. Consistent to previous studies, verapamil at the concentration of 20 µM, which completed abolished 80 mM [K+]o induced vasoconstriction, only attenuated phase I by 19.8 ± 4.5 % (n= 6); whereas SKF 96365 (40 µM) significantly decreased phase I by 31.8 ± 5.6 % (n=6) and 2APB (50 µM) significantly reduced phase I by 18.1 ± 6.8 % (n=7) and phase II by 50.1 ± 7.6% (n=7). These results support that Ca2+ entry via voltage gated channel (L-type) and non-voltage gated channel e.g. TRP channel mediates HPV. In a previous work, we have demonstrated that arachidonic acid (AA) could inhibit hpoxia-induced [Ca2+]i elevation in primary cultured human pulmonary artery smooth muscle cells. However, in rat intrapulmonary artery, AA alone at concentration 10 µM (n=8) induced a significant transient vasoconstriction, equivalent to 29.1 ± 9.8% of vasoconstriction induced by 80 mM [K+]o. AA (10 µM) significantly attenuated the phase II (c) of HPV by 25.5 ± 4.5 % (n=7) and had not effect on phase I of HPV. The inhibitory effect of AA was mediated by downstream metabolites of AA via cylcoxygenase (COX). ETYA is a pharmacology analogue of AA and also inhibits COX, LOX and CYP450 at distinct dosage ranges. ETYA at concentration of 20 µM, which could inhibit COX, significantly inhibited phase I by 59.5 ± 6.5 % and phase II 35.9 ± 7.6 % of HPV (n=9). Whereas ETYA at concentration of 1 µM, which could inhibit LOX, had no effect on either phase I or II of HPV (n=6). Blockage of COX by DuP697 reduced the phase II of HPV by 27.6 ± 6.8 % (n=6) and it was consistent to the effect of AA. Taken together, the data suggested that metabolites of AA via COX involve in mediating phase II of HPV, which is endothelium dependent. Other metabolites in addition to products of COX may participate in mediating phase I of HPV. Blockage of CYP450 by SKF525 significantly attenuate the phase I of HPV by 24.9 ± 7.5 % (n=6). EET (200 nM), a metabolite of AA via CYP450 epoxygenase, reduced the phase I of HPV by 10.5 ± 5.6 % (n=5), suggesting the vasoconstriction might be derived from effects of the metabolites via CYP450 hydroxylase, but not epoxygenase. In addition, either application of PGI2 (n=6) or inhibition of PGI generation by 15 HPETE (n=6) had no significant effects on HPV, implicating HPV in rat intrapulmonary artery is primarily determined by increase of vasoconstrictors and not by reduction of vasodilators.