OBJECTIVE--This study investigated the changes in regional myocardial ultrasonic backscatter, measured as myocardial echo amplitude, that occur during reversible myocardial ischaemia in humans. DESIGN--Left anterior descending coronary angioplasty was used to produce reversible myocardial ischaemia in human subjects. Regional myocardial echo amplitude was studied in the interventricular septum and left ventricular posterior free wall before, during, and after coronary occlusion with the angioplasty balloon. Wall motion analysis of the left ventricle was performed from simultaneous cross sectional echocardiographic imaging. Patients were studied prospectively. PATIENTS--Six patients (mean age 56 (SD 11), range 46 to 69 years) with single vessel, left anterior descending coronary artery stenoses, were investigated during elective coronary angioplasty. A total of 11 balloon inflations were studied. SETTING--All patient studies were performed at Harefield Hospital. Echo amplitude analysis was performed at the Royal Brompton Hospital. INTERVENTIONS--Angioplasty was performed by the usual procedure at Harefield Hospital for elective coronary angioplasty. All routine medication including beta blockers and calcium antagonists were continued. Inflation pressures were up to 12 atm (1212 kPa) and mean inflation time ranged from 30 to 120 (86 (31)) s. In four studies the first inflation was examined, in three the second, in two the third, and in one each the fourth and fifth inflations. Echo amplitude and cross sectional echo-cardiographic studies were recorded with a 3.5 MHz Advanced Technology Laboratories (ATL) (720A/8736 series) mechanical sector scanner and an ATL Mark III (860-1 series) echocardiograph system with 45 dB logarithmic grey scale compression. MAIN OUTCOME MEASURES--Regional echo amplitude was examined in four regions of the left ventricle--namely, the basal and mid-septum, and basal and mid-posterior wall. Consecutive end diastolic and end systolic frames were analysed and cyclic variation was determined as the difference between the level of echo amplitude at end diastole and at end systole. Measurements were made before balloon inflation, at peak inflation, and after balloon deflation. Regional wall motion and systolic wall thickening were analysed qualitatively. RESULTS--Before balloon inflation, cyclic variation in echo amplitude was noted in all regions (basal septum, 2.4 (SD 1.1) dB; mid-septum, 2.5 (1.1) dB; basal posterior wall, 3.3 (2.1) dB; mid-posterior wall, 3.9 (1.6) dB). During balloon inflation there was a significant fall in cyclic variation to 0.4 (0.9) dB (p < 0.0002) in the mid-septum. This was predominantly owing to an increase in end systolic echo amplitude from 5.4 (2.0) dB to 9.3 (1.9) dB (p < or = 0.01). This was associated with the development of severe hypokinesis or akinesis in the mid-septum. No significant changes in echo amplitude occurred in the three other regions examined. Changes were completely reversed after balloon deflation. CONCLUSIONS--These results suggest a causal relation between occlusion of the supplying coronary artery and blunting of myocardial echo amplitude cyclic variation. It is suggested that balloon occlusion produced myocardial ischaemia. The resultant impairment of myocardial contraction then caused a blunting of cyclic variation in echo amplitude. The results of this study provide further data about the ability of quantitative studies of ultrasonic backscatter to identify alterations in the myocardium during injury.