We monitored an index of coronary vascular resistance (mean aortic pressure/mean coronary flow) in 19 heart-blocked conscious dogs paced at 60 beats/min and instrumented with an aortic pressure catheter, left circumflex artery electromagnetic flow probe, and a coronary occluder. Cessation of pacing for a single beat resulted in a long diastole control (LDC) intervention beat of 2-second duration characterized by a progressive rise in diastolic coronary vascular resistance index. A 400-msec coronary artery occlusion early in a long diastole (LD4) dramatically inhibited the rate of rise in resistance index during the first 600 msec (phase 1) after occlusion. Partial recovery of the resistance index rise rate was evident during the remaining 400 msec (phase 2) of the long diastole. In nine dogs, LDC and LD4 intervention beats were instituted during two conditions of myocardial metabolic activity in which the myogenic stimuli associated with coronary occlusion would be similar: 1) paired pacing and 2) normal pacing plus intravenous adenosine and phenylephrine infusions (AP) to maintain mean aortic pressure and coronary flow at paired pacing levels. During paired pacing, the LD4-LDC differences in phase 1 and 2 resistance index rise rates (-69 +/- 18 and -48 +/- 31 mmHg/ml/sec2, respectively) were greater than during normal pacing plus AP (-32 +/- 14 and -1 +/- 32 mm Hg/ml/sec2, phase 1 and 2, respectively) (p less than 0.05). These differences are consistent with operation of a metabolic mechanism in response to occlusion.(ABSTRACT TRUNCATED AT 250 WORDS)