The American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) classification has been used for the diagnosis of breast masses for several decades and constantly updated, but the terminology used to describe breast ultrasound findings is still evolving and a great amount of large sample data is necessary to verify and improve ultrasound BI-RADS. The objective of the present study was to explore the value of ultrasound Breast Imaging Reporting and Data System (US BI-RADS) classification in the preoperative evaluation of the US-guided Mammotome-assisted minimally invasive resection of breast masses. A total of 1,028 patients with 1,341 breast masses from a single hospital were selected for retrospective analysis. All patients underwent minimally invasive resection using a US-guided Mammotome device, and postoperative pathological examinations were performed for all samples. The preoperative US BI-RADS classification and postoperative pathological examination results were compared and analyzed. A receiver operating characteristic (ROC) curve was used to analyze the preoperative evaluation efficacy of the US BI-RADS classification in US-guided Mammotome-assisted minimally invasive breast mass resection. Among the 1,341 breast masses that underwent resection, 1,307 were benign and 34 were malignant. The specificity, sensitivity, accuracy, positive predictive value and negative predictive value of the US BI-RADS classification in the preoperative diagnosis of malignant breast masses were 83.47, 100.00, 83.89, 13.60 and 100.00%, respectively, and the area under the ROC curve was 0.917. It may be concluded that the US BI-RADS classification has a good preoperative diagnostic performance and can provide an accurate assessment prior to Mammotome-assisted minimally invasive resection. It may help surgeons to make reasonable decisions for subsequent therapy and therefore is worthy of further clinical use.