The effects of occlusion, reperfusion, reocclusion (n = 13), and nitroglycerin (n = 10) on regional transmural myocardial collateral blood flow was tested in conscious dogs in which collateral development was stimulated by partial stenosis of the left circumflex coronary artery. Hemodynamics and collateral blood flow were measured during the awake state using 9-micron radioactive microspheres. Regional transmural flow was measured during transient occlusion of the circumflex artery at 7 and at 14 days postoperatively. On the 14th postoperative day, two sets of circumflex occlusions and blood flow measurements were carried out. The first set consisted of two occlusions separated by 15 minutes. The second set performed 2 hours later included two occlusions, separated by 15 minutes, and nitroglycerin administration. Mean collateral blood flow increased significantly (P = 0.002) from 0.10 +/- 0.07 ml/min per g on day 7 to 0.25 +/- 0.18 ml/min per g on day 14. A significant increase in mean collateral blood flow from occlusion one to two was observed (0.28 +/- 0.17 to 0.37 +/- 0.22 ml/min per g, P = 0.005). Mean collateral flow increased significantly (P = 0.01) between pre- to post-nitroglycerin occlusions, 0.28 +/- 0.20 to 0.46 +/- 0.32 ml/min per g. Although this increase appeared to be greater than during the first set of occlusions, it did not reach statistical significance (P = 0.08). These data indicate that when immature collaterals are present, occlusions, reperfusion, and reocclusions of a major coronary artery produce augmentation in collateral flow. This must be considered in evaluating interventions which may alter collateral flow.