The pulmonary vascular responses to 50 per cent nitrous oxide were studied in 32 anesthetized patients ventilated to maintain normal PaCO2. One group consisted of sixteen patients with coronary artery disease (CAD) and normal pulmonary vascular resistance (PVR) about to undergo coronary artery bypass surgery. A second group consisted of 16 patients with markedly elevated PVR values due to chronic mitral valve stenosis (MVS). CAD patients showed a significant increase in PVR irrespective of whether halothane or fentanyl was used as background anesthetic. Individual changes, however, did not exceed the upper limit of normal and therefore are not considered to be of clinical importance in these patients. In patients with MVS subjected to fentanyl anesthesia, N2O caused a marked increase in PVR from 357 to 530 dyn . s. cm-5. Halothane anesthesia did not significantly attenuate the effect of nitrous oxide on the pulmonary vasculature as mean PVR increased from 351 to 451 dyn . s. cm-5. These results suggest that the preexisting PVR value is of more importance for the pulmonary vascular response to N2O than the influence of background anesthesia. We conclude that nitrous oxide should be used with caution in patients with elevated pulmonary vascular resistance, particularly in the presence of right ventricular dysfunction and/or right coronary artery disease.