Abstract Several factors affecting the power of Doppler scattered signal and, consequently, microembolus size estimation, may be eliminated when assessing the microembolus size via multiple measurements. A new microembolus size estimator is proposed based on the ratio of microembolus scattering cross-section in two directions and for two emission frequencies. Theoretical considerations indicate that the estimation of size of microembolic elements should be independent of the spatial distribution of the wave intensity, tissue attenuation and hardware factors. The simulation results indicate that this estimation only slightly depends on the material of the microembolus and acoustic properties of blood. The experimental results indicate that the accuracy of median size estimation increases with microembolus size. The measurement error is less than 27% for microemboli with median diameter larger than 360 μm. The method is constrained to the estimation of microembolus size in the vessels of extremities.