Since the mechanism by which oxygen reduces infarct size is not yet completely understood, we tested the hypothesis that breathing 100% oxygen may alter regional myocardial blood flow. In 23 anaesthetised open-chest dogs with acute occlusions of the left anterior descending coronary artery or of its apical branch, regional myocardial blood flow was determined by the microsphere technique at 15 and 30 min following coronary artery occlusion. Fifteen treated dogs were ventilated with 100% oxygen beginning 17 min after occlusion (ie, after the first determination of regional myocardial blood flow) and continuing until the end of the experiment, while the other 8 dogs served as controls and were ventilated continuously with room air. In treated dogs, following the administration of 100% oxygen, transmural flow at sites remote from the distribution of the occluded vessel (nonischaemic sites) decreased from 121 +/- 5 before oxygen to 108 +/- 6 cm3 . min-1 . 100g-1 (P less than 0.05). However, in the ischaemic sites it increased from 28 +/- 3 before oxygen to 32 +/- 3 cm3 . min-1 . 100g-1 (P less than 0.05). In the control dogs transmural flow, 15 and 30 min after occlusion, was, in nonischaemic sites, 118 +/- 7 and 125 +/- 10 cm3 . min-1 . 100g-1 NS, and in the ischaemic sites 27 +/- 4 and 25 +/- 3 cm3 . min-1 . 100g-1 NS, respectively. Thus, ventilation with 100% oxygen can augment perfusion of the ischaemic myocardium and this may be, in part, the mechanism by which it reduces infarct size.