Isolated guinea pig hearts were used as a model to investigate the role of pO2, coronary perfusion pressure (CPP), and flow on the release of nitric oxide (NO) into the coronary circulation. The in vitro half-life of NO strongly depends on the pO2. Reduction of pO2 from 700 to 50 mm Hg prolonged the NO half-life by 78%, from 3.6 to 6.4 s. Basal release of guansine-3′,5′-cyclic monophosphate (cyclic GMP) taken as an index of endogenous NO concentration, remained unaltered in isolated hearts perfused at normoxic (100% O2) and hypoxic conditions (30% O2). Increase of CPP after infusion of 4 μM HbO2 or 10 μM NG-monomethyl-l-arginine (l-NMMA) (ΔCPP of 3 ± 0.1 and 2 ± 0.1%, respectively) was similiar during normoxic and hypoxic perfusion. Mannitol (25 mM) and dimethyl sulfoxide (10 mM), which are scavengers of hydroxyl radicals, significantly reduced CPP and increased basal cyclic GMP release. Bradykinin (100 nM) decreased CPP by more than 40% at each level studied. Increasing flow from 5 to 15 ml/min enhanced NO release from bradykinin-stimulated hearts from 61 ± 5 to 207 ± 13 pmol/min, determined by difference spectro-photometry. Maximal reduction of CPP by adenosine did not affect basal NO release. The presented data suggest that in vitro the half-life of NO strongly depends on pO2. The NO half-life in situ may be influenced by the formation of hydroxyl radicals within the coronary circulation. Hypoxia-induced coronary vasodilation does not appear to be mediated by NO. NO release from bradykinin-stimulated hearts increases proportionally to enhanced flow rates. Basal release of NO occurs independently from changes in coronary perfusion pressure.