We have recently shown that the cardiac cycle-dependent variation in myocardial ultrasonic integrated backscatter is blunted with regional ischemia in dogs. To determine if global and intramural regional myocardial contractile performance can be quantified by integrated backscatter, we analyzed ultrasonic responses after induction of increased and decreased contractility in five dogs. A recently developed analog data-acquisition system for measuring integrated backscatter in real time was used to sample radiofrequency signals gated from subepicardial or subendocardial regions. Base-line recordings of integrated backscatter, left ventricular pressure, left ventricular dP/dt, and wall thickness were made at 12 left ventricular sites for both intramural regions. Contractility was modified subsequently by either paired pacing or propranolol to produce significantly elevated or depressed values for maximum left ventricular dP/dt compared with baseline (1083 +/- 289 to 3001 +/- 570 mm Hg/sec; p less than .01 for all). The amplitude of the cyclic variation of integrated backscatter was 50% greater (arithmetically) in subendocardial than in subepicardial regions for all treatments (7.6 +/- 0.3 vs 6.0 +/- 0.5 dB, p less than .001). The maximum rate of change in integrated backscatter waveforms during isovolumetric contraction was faster with paired pacing and slower with propranolol than at baseline for all regions (56 +/- 6 to 74 +/- 6 to 82 +/- 5 dB/sec, p less than .005). The maximum rate of change in integrated backscatter also was greater in subendocardial than subepicardial regions (p less than .001). Thus, both regional and global differences in myocardial contractile performance are manifest quantitatively in integrated backscatter waveforms. We propose that the physiologic determinants of these differences may depend on regional and global variations in myofibril elastic characteristics.