RATIONALE AND OBJECTIVES. TO study the relative contributions of different bubble size classes to SonoVue(TM)'s echogenicity in fundamental acoustic imaging modes. SonoVue(TM) is a contrast agent, previously known as BR1, with a bubble size distribution extending from approximately 0.7 to 10 mum, METHODS. A model for the acoustic response of SonoVue(TM) was determined and validated for a set of experimental data. This model was used to simulate the acoustic response of a standard batch of SonoVue(TM) as the sum of responses of nonoverlapping bubble size classes. RESULTS. The simulation was first validated for a standard SonoVue(TM) bubble size distribution. When this distribution was considered as five size classes with equal numbers of bubbles, it was found that bubbles smaller than 2 mum accounted for 60% of the total number but contained only 5% of the total gas volume. The simulation results indicated marked differences in the acoustic contributions from these classes, with 80% of the acoustic efficacy provided by bubbles 3 to 9 mum in diameter. The study also compared bubble distributions in number, surface, and volume, with the distribution computed in terms of acoustic efficacy, CONCLUSIONS. This study shows why bubble volume is a much better indicator of SonoVue(TM)'s efficacy than is bubble count. A low threshold in diameter was found for SonoVue(TM) microbubbles at approximately 2 mum, under which size bubbles do not contribute appreciably to the echogenicity at medical ultrasound frequencies.