In 10 conscious dogs, a model was developed for studying regional contractile responses in a coronary collateral-dependent bed. Regional myocardial function was compared after terminating a maximum paced rate of 240 beats/min maintained for 3 minutes (postpacing period) with that during telemetry-monitored exercise at comparable heart rates (average 252 +/- 34 beats/min, duration 2.4 minutes) at different times during collateral development. Ultrasonic dimension gauges were used to measure control and ischemic segment (CS and IS) lengths and ischemic zone regional wall thickness (IW). An ameroid constrictor and a Doppler flow probe were placed around the left circumflex coronary artery, and pacing electrodes were sutured to the right ventricle. An average of 23 days postoperatively, coronary obstruction was complete. Studies at that time showed that percent shortening (% delta L) of IS and percent wall thickening (% delta W) of IW decreased after pacing to 57% and 35% of control, respectively, and during exercise to 37% of control. One week later (average 30 days postoperatively), significant depression of regional function no longer occurred postpacing. However, exercise at a comparable heart rate still provoked regional dysfunction in the collateral-dependent zone: Both IS% delta L and IW% delta W decreased to 51% of control. Regional function at rest did not differ during these studies. Thus, the effectiveness of the postpacing response for detecting limited collateral reserve was eliminated by further collateral development, but regional myocardial dysfunction during exercise stress served to detect ischemia despite increased collateral circulation.