In patients undergoing aortic valve surgery during cardioplegia (a condition characterized by extra-corporal circulation, potassium-induced cardiac arrest and cardiac hypothermia) cardiac release or uptake of 6-keto-PGF 1α, and in some cases also of 6,15-diketo-13,14-dihydro-PGF 1α were determined, using radioimmunoassay with specific antibodies for determination of the arterial and coronary sinus plasma levels of these PGI 2 metabolites. The aim of the study was to find out whether cardiac hypoxia, as evidenced by anaerobic glycolysis in the heart, resulted in stimulation of it PGI 2 production. Before cardioplegia the uptake of oxygen (23 ml/min) and the arterio-coronary sinus concentration difference (a-cs) of lactate (0.3 mM) were within the normal range, indicating that the O 2 supply to the heart was sufficient for aerobic metabolism. During cardioplegia the oxygen uptake fell, to less than 3 % of the basal value. The uptake of lactate was significantly reversed into a small release inicating that myocardial hypoxia developed. No release of 6-keto-PGF 1α took place before cardioplegia, implying that the hearts' PGI 2 productino was low or absent when its O 2 supply was sufficient. During cardioplegia a significant (a-cs) of 6-keto-PGF 1α developed, amounting to −75 pg/ml. Since no evidence of a hypoxia-induced decrease in the metabolic breakdown of PGI 2 or 6-keto-PGF 1α in the heart could be demonstrated - as shown by a continued, or even augmented liberation of 6,15-diketo-13,14-dihydro-PGF 1α during cardioplegia - the cardiac release of 6-keto-PGF 1α indicated that the PGI 2 production in the heart was stimulated during the cardioplegia/hypoxia. From these data we suggest that human cardiac formation of PGI 2 is low or absent when the O 2 supply is sufficient. Furthermore, cardiac hypoxia, as evidenced by release of lactate, results in stimulation of cardiac PGI 2 formation.