OBJECTIVES: To develop a predictive model using conventional ultrasound combined with CEUS to identify thoracic wall recurrence after mastectomy. METHODS: A total of 162 women with pathologically confirmed thoracic wall lesions (benign 79, malignant 83; median size 1.9 cm, ranging 0.3-8.0 cm) underwent a mastectomy and were checked by both conventional ultrasound and CEUS and were retrospectively included. Logistic regression models of B-mode ultrasound (US) and color Doppler flow imaging (CDFI) with or without CEUS were established to assess the thoracic wall recurrence after mastectomy. The established models were validated by bootstrap resampling. The models were evaluated using calibration curve. The clinical benefit of models were assessed using decision curve analysis. RESULTS: The area under the receiver characteristic was 0.823 (95% CI: 0.76, 0.88) for model using US alone, 0.898 (95% CI: 0.84, 0.94) for model using US combined with CDFI, and 0.959 (95% CI: 0.92, 0.98) for model using US combined with both CDFI and CEUS. The diagnostic performance of the US combined with CDFI was significantly higher than that of the US alone (0.823 vs 0.898, p = 0.002) but significantly lower than that of the US combined with both CDFI and CEUS (0.959 vs 0.898, p < 0.001). Moreover, the unnecessary biopsy rate of the US combined with both CDFI and CEUS was significantly lower than that of the US combined with CDFI (p = 0.037). CONCLUSIONS: Compared to B-mode ultrasound and CDFI, CEUS improves the diagnostic performance to evaluate thoracic wall recurrence after mastectomy. KEY POINTS: • CUES is an effective supplementary method for US in the diagnosis of thoracic wall recurrence after mastectomy. • CEUS combined with both US and CDFI can significantly improve the accuracy of diagnosis of thoracic wall recurrence after mastectomy. • CEUS combined with both US and CDFI can reduce the rate of unnecessary biopsy of thoracic wall lesions after mastectomy.