BACKGROUND: It is generally accepted that propofol does not inhibit hypoxic pulmonary vasoconstriction (HPV). However, because the previous studies for the effects of propofol on HPV were established in vivo, the effects of physiologic variables could not be ruled out. Therefore, we investigated the effects of various concentrations of propofol on HPV at isolated rat lungs and the relationship of these effects of propofol on HPV and endothelium-derived relaxing factor (EDRF) and an ATP-dependent K+ channel which were candidates as the mechanism of HPV. METHODS: In 30 isolated rat lungs, after three hypoxic challenges for 5 minutes, we administered saline in the control group, N(G)-nitro-L-arginine methyl ester (L-NAME) in the L group and glibenclamide in the G group followed by three hypoxic challenges for 5 minutes. In addition, we studied the effects of various concentrations of propofol on HPV in the three groups. RESULTS: L-NAME and glibenclamide did not alter baseline pulmonary arterial pressure but L-NAME significantly enhanced HPV. Clinical concentrations of propofol did not affect HPV and high concentrations of propofol inhibited HPV. The pretreatment of L-NAME and glibenclamide did not alter the inhibition of HPV even at high concentrations of propofol. CONCLUSIONS: The EDRF and ATP-dependent K+ channel did not largely contribute to baseline pulmonary arterial tone but EDRF might be released and downregulate HPV. Clinical concentrations of propofol did not inhibit HPV but high concentrations of propofol inhibited HPV. In addition, the mechanism of inhibition of HPV at high concentrations of propofol did not relate to the EDRF pathway and ATP-dependent K+ channel.