summaryDilator products of nitric oxide synthase (NOS) and cyclooxygenase (COX) may contribute to the low normal pulmonary artery pressure (Ppa). In isolated perfused lungs of ferrets, rabbits and rats we investigated this hypothesis by blockade of NOS with ¬_NAME (l-nitro-arginine methyl ester) and COX with meclofenamate. There were species differences. Inhibition of either enzyme caused little rise in Ppa in ferrets and rats but inhibition of both enzymes caused huge increases in Ppa. We suggest this might be due to intracellular connections between the excitatory pathways for NOS and COX dilators, such that inhibition of one enzyme leads to activation of the other. Impairment of these endothelial-based enzymes in pulmonary vascular disease might lead to severe pulmonary hypertension. By contrast, in rabbits, comparable doses of ¬_NAME lead to large rises in Ppa which were reversed rather than amplified by COX blockade. NO seems to protect against some pressorÏoedema forming product of COX in this species. introduction Nitric oxide synthase (NOS) and cyclooxygenase (COX) are present in pulmonary vascular endothelium and their dilator products nitric oxide (NO) and prostacyclin (PGIµ) contribute to pulmonary dilatation. The circumstances which cause their release are not fully worked out and may differ with species. In isolated rat lungs perfused at constant flow, blockade of either NOS with the ¬_arginine analogues (Mazmanian et al. 1989;Archer et al. 1990;Hasunuma et al. 1991;Barer et al. 1993) or COX with sodium meclofenamate (Russell et al. 1993) caused little or no rise in pulmonary artery pressure (Ppa); however, if the vasculature was in a state of vasoconstriction caused by vasoconstrictor agents or chronic exposure to hypoxia, ¬_NAME did cause a substantial rise in Ppa. These results suggest that, in these circumstances, there is no constant release of NO or dilator prostaglandins in the rat. Cremona et al. (1994) who used a similar preparation, also found that ¬_NAME did not alter pulmonary vascular resistance (PVR) during normoxia in dog lungs but raised it in the lungs of sheep, pigs and humans; in all four species, ¬_NAME raised pulmonary vascular resistance during hypoxic vasoconstriction (HPV). In rabbits, in vivo, Persson et al. (1990) and, in intact cats Hyman et al. McMahon et al. (1991) found that NOS blockade raised Ppa during normoxia. However, in conscious dogs, Nishiwaki et al. (1992) found no evidence that NO was released in the absence of experimental stimuli. The work of Barnard et al. (1993) suggested that either NOS or COX may be the predominant pulmonary dilator enzyme in different species. They used isolated lungs perfused at constant pressure and found that NOS but not COX