Objective To explore the classification of ultrasound breast imaging reporting and data system (BI-RADS) classification combined with serum trefoil factor 1 (TFF1) and human growth differentiation factor 3 (GDF3) in the differential diagnosis of benign and malignant breast masses. Methods The prospective study collected 113 female patients with breast mass who got treatment in Tangshan People’s Hospital from September 2020 to September 2023. Ultrasound diagnostic equipment was applied for ultrasound BI-RADS classification, ELISA method was applied to detect serum TFF1 and GDF3 levels, Consistency Kappa test was used to compare the consistency of ultrasound BI-RADS classification, serum TFF1 and GDF3 alone and combined in the diagnosis of benign and malignant breast masses and pathological results, In addition, receiver operating characteristic (ROC) curve was used to analyze the diagnostic value of the above indicators in the diagnosis of benign and malignant breast masses. Results Among the 113 patients with breast mass, the pathological diagnosis showed benign mass in 48 cases and malignant mass in 65 cases. The serum levels of TFF1 [(1.62±0.25) ng/mL vs (1.24±0.27) ng/mL] and GDF3 [(118.62±21.73) pg/mL vs (96.47±16.05) pg/mL] in malignant patients were obviously higher than those in benign patients (P<0.001). There were 70 cases of malignant breast mass and 43 cases of benign breast mass according ultrasound BI-RADS classification combined with serum TFF1 and serum GDF3. The diagnosis consistency with pathological results was very high and Kappa value was 0.835 (P<0.001). The sensitivity and specificity of ultrasound BI-RADS classification combined with serum TFF1 and serum GDF3 in differential diagnosis of benign and malignant breast masses were 96.92% (63/65), 85.42% (41/48), respectively, the negative predictive value was 95.35 (41/43), and positive predictive value was 90.00% (63/70). The sensitivity, negative predictive value and accuracy were higher than those of ultrasonic BI-RADS classification and serum GDF3 alone (P<0.05), and the missed diagnosis rate was lower than those of ultrasonic BI-RADS classification and serum GDF3 alone (P<0.05). Conclusions The serum levels of TFF1 and GDF3 increase in patients with malignant breast masses. The combination of ultrasound BI-RADS classification and serum TFF1 and GDF3 can improve the sensitivity and accuracy of the diagnosis of benign and malignant breast masses, and reduce the missed diagnosis rate.