PURPOSE: Accurate identification and assessment of sentinel lymph node (SLN) using noninvasive imaging methods can play a vital role in tumor staging, surgical planning, and prognostic evaluation. In this study, we assessed the efficacy of B7-H3-targeted molecular-ultrasound imaging for the early SLN detection, and characterization in a mouse model of orthotopic breast cancer. PROCEDURES: We established a mouse breast cancer model with lymph node metastasis by injecting MAD-MB 231 cells which were engineered to express firefly luciferase reporter gene into the fat pad of the right 4th mammary gland in female BALB/c nude mice. The sole lymph node (LN) close to the tumor was regarded as the SLN for imaging investigation, which included metastatic and non-metastatic SLNs. The LN in the right 4th mammary gland from normal mice was used as normal control (normal mice LN). The commercially available preclinical streptavidin-coated, perfluorocarbon-containing lipid-shelled microbubbles (VisualSonics, Toronto, Canada) were used to generate B7-H3-targeted microbubbles (MB) and control microbubbles (MB). Then, ultrasound molecular imaging (USMI) was performed using a high-resolution transducer (MS250; center frequency, 21 MHz; Vevo 2100; VisualSonics, Toronto, Canada) after intravenous injection of microbubbles. RESULTS: The SLN was clearly detected and located under conventional (B-mode) and contrast-enhanced ultrasonography with microbubble injection. The metastatic SLNs showed a markedly higher signal from B7-H3-targeted microbubbles (MB) compared to the non-metastatic SLNs and normal LNs. The metastatic SLN was further confirmed by ex vivo bioluminescence imaging and eventually verified by histological analysis. CONCLUSIONS: Our findings suggest the potential value of USMI using B7-H3 targeted microbubbles in breast cancer and establish an effective imaging method for the non-invasive detection and characterization of SLN.