The possible role of sarcoplasmic reticulum (SR) Ca2+ stores in hypoxic pulmonary vasoconstriction (HPV) is not well understood. In order to assess the possible role of intracellular Ca2+ release from SR Ca2+ stores in HPV, we examined the effects of: (1) ryanodine (10 microM) depletion of intracellular Ca2+ stores, and (2) thapsigargin (THAPS, 2 microM) or cyclopiazonic acid (CPA, 10 microM) depletion of intracellular Ca2+ stores on HPV in canine pulmonary artery. 2 Isometric tension was measured from arterial ring suspended in Krebs-Henseliet solution (K-H) bubbled with 95%O2/5%CO2. Hypoxia was induced by bubbling phenylephrine (PE, 1 microM) precontracted rings with 95%N2/5%CO2. HPV was observed in both intact and endothelial-denuded arteries and expressed as % of maximal KCl contraction (% Tkmax) = 21.3 +/- 3.2%; n = 13 and 21.7 +/- 4%; n = 4 respectively. 3 When SR caffeine sensitive Ca2+ stores were depleted by pretreatment with ryanodine and brief caffeine (15 mM) exposure, the hypoxic response was significantly reduced to 19.1 +/- 9.2% of the control hypoxic contraction (n = 7; p < 0.001) with little or no effect on PE or KCl contractions. On the other hand, in normoxic rings pretreated with THAPS or CPA, the PE responses were significantly reduced (% Tkmax = 18.2 +/- 3.1% compared to 39.0 +/- 3.9% in control; n = 16; P < 0.001; %Tkmax = 3.4 +/- 1.6% compared to 49.9 +/- 7.9% in control; n = 6; P < 0.001; respectively) with no significant effect on caffeine-induced contractions, suggesting that both THAPS and CPA preferentially deplete InsP3-sensitive Ca2+ stores, without affecting the caffeine-sensitive Ca2+ store; consistent with the existence of separate and independent InsP3 and caffeine-sensitive Ca2+ stores in this preparation. 4 When hypoxia was induced in the presence of THAPS or CPA, developed tension was significantly larger than control (% Tkmax = 64.5 +/- 6.0%; n = 16; P < 0.05%; %Tkmax = 78.2 +/- 15%; n = 6; P < 0.05; respectively), was partially blocked by nisoldipine (10 microM) and ryanodine (% Tkmax = 20.3 +/- 3.7%; n = 6), and nearly completely blocked by SK&F 96365 (50 microM). However, the actions of SK&F 96365 appeared to be nonselective since this compound also significantly reduced contractions elicited by KCl, PE and caffeine. 5 Finally, evidence was obtained suggesting: (a) that at least some of the Ca2+ released from the caffeine- and ryanodine-sensitive Ca2+ stores by hypoxia may be taken up and buffered by the InsP3-sensitive Ca2+ stores, and (b) the apparent dependence of HPV on extracellular Ca2+ entry pathways may be partially due to the dependence of the Ca2+ content of intracellular SR Ca2+ stores on sarcolemmal Ca2+ entry pathways. 6 These data suggest that caffeine- and ryanodine-sensitive SR Ca2+ stores contribute significantly to HPV under normal conditions and, in the presence of THAPS or CPA, an additional nisoldipine- and ryanodine-insensitive Ca2+ entry pathway is evoked by hypoxia.