Breast ultrasound is often used as a diagnostic tool to diagnose breast cancer, however the high false positive rate limits its use in screening. Coherence-based beamforming has been shown improve the distinction between solid and fluid breast masses, therefore the objective of this work is to both qualitatively and quantitatively investigate the clinical impact of coherence-based beamforming. Five board-certified breast radiologists were asked to read ultrasound images of twentysix masses and select the content (i.e., solid, fluid, mixed, or uncertain) and the clinical diagnosis (i.e., BI-RADS 2, 3, 4 or 5). The responses were compared with and without the inclusion of coherence-based beamforming to qualitatively assess the impact of coherence-based beamforming. In addition, coherence-based metrics including lag-one coherence (LOC) and coherence length (CL) were used to quantitatively distinguish solid from fluid-filled masses. When including coherence-based beamforming, the mean reader sensitivity for detection of fluid-filled masses was improved from 57% with B-mode alone to 86% with the addition of coherence-based images. Using LOC as a quantitative metric, with an optimal threshold of 0.3, the sensitivity for detection of fluid-filled masses was further improved to 100% with a specificity of 94%. These results are promising for the inclusion of coherence-based features in the breast clinic in order to improve diagnostic certainty, particularly when distinguishing between solid and fluid-filled breast masses.