The hypotheses tested were that sonolysis of erythrocytes in the presence of a gas-based ultrasound contrast agent in vitro will be related quantitatively to the duration and number of ultrasound pulses applied using a constant pulse repetition period and, at least qualitatively, to the total exposure duration (i.e., the product of pulse number x pulse duration). An objective was to determine the influence of sample rotation during insonation on the amount of hemolysis produced under these conditions, Human erythrocytes, suspended to 40% hematocrit in autologous plasma containing 3.6% (V:V) Albunex(R), were exposed/sham-exposed to 1-100 pulses of 1-MHz ultrasound (6.2 MPa peak positive, 3.6 MPa peak negative acoustic pressures; I-SPTP similar to 800 W/cm(2)) using a 1-s pulse repetition period. Pulse durations ranged from 20-20,000 mu s; samples were either stationary or rotated (200 rpm) during insonation, Hemolysis was independent of vessel rotation treatment at all tested pulse durations and pulse numbers. Levels of hemolysis statistically greater than in sham-exposed samples were obtained with greater than or equal to 50 pulses of 20 mu s duration, and greater than or equal to 1 pulse of 200, 2000 or 20,000 mu s duration. Hemolysis increased with increasing pulse number and poise duration. Approximately equivalent levels of hemolysis were produced by different pulse number x pulse duration combinations, yielding the same total exposure duration. (C) 1999 World Federation for Ultrasound in Medicine & Biology.