This study tests the hypothesis that acute myocardial ischemia induces a characteristic temporal variation in regional ultrasound amplitudes. Myocardial ischemia was created by circumflex coronary artery occlusion in seven closed-chest mongrel dogs. Ultrasound images were acquired prior to occlusion and post-occlusion on a phased-array two-dimensional system. Unprocessed ultrasound data from end-diastolic images were taken digitally for quantitative gray level analysis. Temporal variation in ultrasonic gray level of a nonischemic control region was compared to the temporal variation in gray level of the ischemic area. In the ischemic area, the average gray level for all seven dogs increased from 39.2 +/- 4.2 prior to occlusion to 42.5 +/- 4.9 at 15 minutes after occlusion, and then to 44.4 +/- 5.9, 45.3 +/- 6.2, and 47.0 +/- 6.0 at 30, 60, and 120 minutes, respectively (p less than 0.05 for control vs 15 minutes and 15 minutes vs 120 minutes). No significant changes in the average gray level of all seven dogs occurred in the nonischemic area from pre-occlusion to 2 hours post-occlusion (38.8 +/- 8.8, 38.4 +/- 8.0, 37.7 +/- 8.4, 37.8 +/- 8.5, and 38.0 +/- 8.2 for control, 15, 30, 60, and 120 minutes, respectively. These data show that regions of acute myocardial ischemia can be characterized by temporal variation in intramyocardial ultrasonic gray level, not only from the time before coronary occlusion to 15 minutes after occlusion, but also between 15 and 120 minutes in the post-occlusion period. Gray level values in nonischemic regions of the left ventricle are remarkably constant over time.