There is continuing controversy over the effects of nitroglycerin and nitroprusside on myocardial ischemia. In 36 open-chest, anesthetized dogs with normal left ventricular filling pressures, intramural carbon dioxide tension (PmCO2) was measured directly with a mass spectrometer during repeated 10-minute coronary artery occlusions separated by 45-minute periods of reflow. Simultaneously, regional myocardial blood flow (RMBF) in the ischemic area was quantified by the microsphere technique. In all dogs the increase in PmCO2 from before to 10 minutes after the first occlusion (delta PmCO2) exceeded that during subsequent occlusions. In those dogs not receiving an intervention (controls), delta PmCO2 during the third occlusion was similar to that during the second occlusion. When nitroglycerin was administered before the third occlusion, it caused a significantly smaller elevation in delta PmCO2 than that which occurred during the control second occlusion, and transmural RMBF to the ischemic region was not altered. In contrast, sodium nitroprusside sufficient to reduce mean systemic arterial pressure to the same extent as nitroglycerin caused no change in delta PmCO2 and a significant decrease of transmural RMBF to the ischemic region. Thus, in the dog with normal left ventricular filling pressures, nitroglycerin reduces myocardial carbon dioxide tension, but nitroprusside given to produce similar hemodynamic alterations exerts not effect on intramural PCO2.