Both the isolated perfused rabbit heart and kidney are capable of synthesizing prostaglandin (PG) I(2). The evidence that supports this finding includes: (a) radiochemical identification of the stable end-product of PGI(2), 6-keto-PGF(1alpha), in the venous effluent after arachidonic acid administration; (b) biological identification of the labile product in the venous effluents which causes relaxation of the bovine coronary artery assay tissue and inhibition of platelet aggregation; and (c) confirmation that arachidonic acid and its endoperoxide PGH(2), but not dihomo-gamma-linolenic acid and its endoperoxide PGH(1), serve as the precursor for the coronary vasodilator and the inhibitor of platelet aggregation. The rabbit heart and kidney are both capable of converting exogenous arachidonate into PGI(2) but the normal perfused rabbit kidney apparently primarily converts endogenous arachidonate (e.g., generated by stimulation with bradykinin, angiotensin, ATP, or ischemia) into PGE(2); while the heart converts endogenous arachidonate primarily into PGI(2). Indomethacin inhibition of the cyclo-oxygenase unmasks the continuous basal synthesis of PGI(2) by the heart, and of PGE(2) by the kidney. Cardiac PGI(2) administration causes a sharp transient reduction in coronary perfusion pressure, whereas the intracardiac injection of the PGH(2) causes an increase in coronary resistance without apparent cardiac conversion to PGI(2). The perfused heart rapidly degrades most of the exogenous endoperoxide probably into PGE(2), while exogenous PGI(2) traverses the heart without being metabolized. The coronary vasoconstriction produced by PGH(2) in the normal perfused rabbit heart suggests that the endoperoxide did not reach the PGI(2) synthetase, whereas the more lipid soluble precursor arachidonic acid (exogenous or endogenous) penetrated to the cyclooxygenase, which apparently is tightly coupled to the PGI(2) synthetase.