Objective To explore the diagnostic value of ultrasound elastography in breast imaging reporting and data system (BI-RADS) category 4 breast nodules. Methods Ninety patients with BI-RADS category 4 breast nodules were enrolled. Preoperatively, all patients underwent conventional ultrasound, strain elastography (SE) and shear wave elastography (SWE). The SE included elastic score, strain ratio (SR) and elastogram-to-B-mode (E/B) ratio, while the SWE covered the maximum elastic modulus (Emax) within the nodule and its surrounding approximately 2 mm shell region (shell 2 mm Emax) , as well as stiff rim sign. Taking histopathological results as the gold standard, receiver operating characteristic (ROC) curves were plotted to compare the diagnostic performance of conventional ultrasound BI-RADS classification, SE, SWE, and combined application of multiple ultrasound techniques for BI-RADS category 4 breast nodules. The conventional ultrasound BI-RADS classification was corrected according to the optimal cut-off values of elastography parameters, and the diagnostic performance of breast nodules was evaluated on the basis of the new corrected classification. Results Of the 90 patients with BI-RADS category 4 breast nodules, 52 had malignant lesions and 38 had benign lesions. Taking pathological results as the gold standard, the malignancy rates of category 4a, 4b and 4c nodules were 6.90%, 57.69% and 100.00%, respectively. The optimal indicator for diagnostic efficacy of SE was SR, and the optimal indicator of SWE was the shell 2 mm Emax. The sensitivity, specificity and area under the curve (AUC) of conventional ultrasound BI-RADS classification for differentiating benign and malignant breast nodules were 0.962, 0.711 and 0.836, respectively. Those of SE combined with conventional ultrasound BI-RADS classification were 0.942, 0.974 and 0.985. Those of SWE combined with conventional ultrasound BI-RADS classification were 0.981, 0.711 and 0.846. Those of conventional ultrasound BI-RADS classification combined with SWE and SE were 0.942, 0.947 and 0.985. After classification adjustment based on ultrasound elastography parameters, the sensitivity, specificity and AUC of the corrected BI-RADS classification were 0.846, 1.000 and 0.962. Its AUC, specificity and sensitivity were all improved compared with the BI-RADS classification. Conclusion SE and/or SWE combined with conventional ultrasound BI-RADS classification can improve the diagnostic performance for benign and malignant BI-RADS category 4 breast nodules and reduce unnecessary puncture biopsy.