Purpose. To identify probably benign breast masses using gray-scale sonography and to see if this strategy could reduce the number of biopsies of breast masses.Methods. This retrospective study included 229 masses in 203 women who underwent sonographically guided percutaneous biopsy. Masses with a negative predictive value for malignancy >98% were retrospectively considered probably benign, and the potential impact of gray-scale sonography in reducing the number of biopsies if these masses were not biopsied was assessed. Assessments were performed considering all masses as a group as well as various subgroups.Results. Round, ellipsoid, or lobulated masses with 3 or fewer lobulations, circumscribed margins, a longitudinal-anteroposterior diameter ratio 1.0 and no marked hypoechogenicity, posterior acoustic shadowing, internal microcalcifications, or altered surrounding breast tissue were considered probably benign. The sensitivity of gray-scale sonography to identify this subgroup was 98%, with a negative predictive value of 99%. If these masses were not biopsied, there would be a 42% reduction in the number of biopsies considering all masses, a 36% reduction for masses classified as Breast Imaging Reporting and Data System category 4, and a 59% reduction for masses exclusively analyzed with sonography.Conclusions. It is possible to identify probably benign breast masses using gray-scale sonography, and thereby to reduce the number of biopsies performed.