Background <p>This study aimed to determine whether contrast-enhanced ultrasound (CEUS) gradient perfusion features in identifying sentinel lymph node (SLN) metastasis burden in early breast cancer (EBC).</p> Materials and methods <p>Data from 145 EBC patients who underwent preoperative CEUS and SLN biopsy were retrospectively collected. The initial time-intensity curve (TIC) parameters and gradient TIC parameters of the high-perfusion regions at the margins and low-perfusion regions inside SLNs in both positive and negative SLN lesions were analyzed to identify independent factors influencing SLN status, and to explore their potential relationship with the average SLN positivity rate and the number of positive SLNs.</p> Results <p>A total of 113 patients were included in the analysis, with 79 in the training set and 34 in the testing set. Significant differences were observed in the gradient TIC parameters, including gradient time-to-peak (△TTP) (<i>t</i> = 3.158, <i>P</i> = 0.002), gradient peak intensity (△PI) (<i>t</i> = 3.934, <i>P</i>&lt;0.001), gradient rising slope (△RS) (<i>t</i> = 2.879, <i>P</i> = 0.005), and gradient area under the curve (△AUC) (<i>t</i> = 5.606, <i>P</i>&lt;0.001) between the two groups. Multivariate logistic regression analysis showed that △PI (<i>P</i> = 0.009), △RS (<i>P</i> = 0.025), and △AUC (<i>P</i>&lt;0.001) were independent factors influencing SLN status and were closely related to the average SLN positivity rate (<i>r</i>_△PI=0.579; <i>r</i>_△RS=0.624; <i>r</i>_△AUC=0.698) and the number of positive SLNs (△PI: <i>P</i> = 0.029; △RS: <i>P</i> = 0.005). Moreover, the SLN status identification model and the classification model predicting whether the number of positive SLNs exceeded two, based on gradient TIC parameters, demonstrated good calibration and clinical applicability in both the training and testing sets.</p> Conclusion <p>CEUS gradient perfusion features can effectively identify SLN metastasis in EBC and are closely related to the number of positive SLNs.</p>