The influence of isoflurane and adenosine on left ventricular myocardial blood flow was investigated in dogs chronically instrumented for measurement of systemic and coronary hemodynamics, regional myocardial contractile function (with ultrasonic sonomicrometers), and myocardial perfusion (by the radioactive microsphere method). An Ameroid constrictor was implanted on the left circumflex coronary artery to produce a progressive stenosis that gradually reduced vascular reserve of the distal perfusion territory. The depletion of reserve was evaluated by daily monitoring of the hyperemic response to adenosine. A stenosis of moderate severity was considered present when left circumflex reserve was attenuated by approximately 60-70%. During left circumflex stenosis development, the left anterior descending coronary artery was totally occluded for 2 min each hour eight times daily with a hydraulic occluder to stimulate coronary collateral development over a period of 9-13 days. Contractile dysfunction during and flow debt repayment after each brief occlusion were used to monitor coronary collateral development. After stenosis and collateral development had occurred, the left anterior descending coronary artery was permanently occluded to simulate a condition of multivessel coronary artery disease with enhanced collateral development. In separate groups of experiments, hemodynamics and myocardial perfusion were measured before and after administration of adenosine (0.54 and 1.08 mg/min) or isoflurane (1.1 and 1.9%, end-tidal) and in the presence of either agent during adjustment of diastolic aortic pressure and heart rate to control levels. Total left anterior descending coronary artery occlusion in the presence of a moderate left circumflex stenosis produced an increase in mean arterial and left ventricular end diastolic pressures. Isoflurane decreased arterial pressure, left ventricular systolic pressure, and positive rate of increase of left ventricular pressure (dP/dt50) without altering heart rate. Administration of the high concentration of isoflurane reduced blood flow in normal areas and in regions distal to the partial (from 1.05 +/- 0.10 to 0.76 +/- 0.11 ml.min-1.g-1) or total coronary occlusion (from 0.64 +/- 0.10 to 0.41 +/- 0.11 ml.min-1.g-1). However, when arterial pressure and heart rate were restored to levels present in the conscious state, perfusion in all zones was maintained at control levels (1.06 +/- 0.11 for the stenotic and 0.69 +/- 0.12 ml.min-1.g-1 for the occluded region). Ratios of transmural blood flow between occluded and normal or occluded and stenotic zones were not different from the conscious state during a constant aortic pressure and heart rate.(ABSTRACT TRUNCATED AT 400 WORDS)