To determine whether contraction and relaxation influence quantitative myocardial ultrasonic backscatter we measured systolic and diastolic integrated backscatter separately in 10 pentobarbital-anesthetized dogs with defined, paced heart rates, before and after coronary occlusion. Data were acquired from intramural sites by coupling a broadband 5-MHz transducer to the left ventricular epicardium. Integrated backscatter was obtained from seven sequential ECG gated intervals throughout the cardiac cycle over the frequencies of 2.5-7.5 MHz and referenced to values obtained with a steel reflector. Before coronary occlusion myocardium in all dogs exhibited a decrease in integrated backscatter from end diastole to end systole (P less than 0.05) in control zones and in zones destined to become ischemic (P less than 0.05). Thirty minutes after occlusion integrated backscatter did not change in control zones but was elevated in ischemic zones with blunting of the diastolic-to-systolic variation. Thus myocardium undergoing contraction exhibits a decrease in integrated backscatter, and measurement of integrated backscatter at end systole differentiates ischemic from normal myocardium.