Many investigators have attempted to use ultrasound imaging to differentiate benign from malignant solid breast masses. Studies have evaluated several generations of gray-scale imaging, Doppler, color Doppler, and power Doppler imaging, and several unconventional ultrasound techniques. Although various individual studies have shown promise, ultrasound criteria for avoiding biopsy of solid lesions have not been widely adopted. Considerable observer variability also remains an important obstacle. This article reviews the results, strengths, and weaknesses of some of the key studies addressing this issue. Fundamental criteria for a successful ultrasound model are also specified.