Emerging evidence indicates that the host microbiota affect tumor progression across various cancers, that microbiota are components of tumors that were once thought to be entirely sterile, and that bacterial populations associated with primary tumors maintain that association at distal metastatic sites. In addition, the introduction of bacteria into tumors can promote their progression. Fu et al. interrogated the association of tumor cells and intratumoral bacteria in a mouse model of spontaneous breast cancer metastasis to better recapitulate human cancer progression. Mouse breast tissue was enriched with microbial communities consistent with human breast tissue, and bacteria lived within the cytosol of the breast tumor cells. Most of the microbes associated with tumors were intracellular. The authors established an antibody treatment approach to target both gut and tumor microbiota or tumor microbiota alone. Eliminating tumor microbiota did not affect primary tumor growth but suppressed metastatic node formation in the lungs. Mice lacking detectable gut microbiota were grafted with bacteria-containing tumors and given water or antibiotics; as before, tumor weights were not different, but metastasis was suppressed in mice that received antibiotics. Because tumor-associated bacteria were involved in the promotion of metastasis, circulating tumor cells (CTCs) were analyzed. Clustered CTCs were more enriched with bacteria compared with primary tumors, suggesting a benefit conferred by the presence of bacteria in CTCs. Bacteria were traced with an erythromycin resistance element–GFP cassette inserted in the Staphylococcus xylosus genome. When injected into tumors, bacteria were detected in lung metastatic outgrowths, but no bacteria were detected in lungs without metastases. When injected into tail veins, bacteria were detected in lungs only when tumor cells were also present, suggesting that the bacteria required tumor cells for colonization. The authors further determined that bacteria also triggered metastasis when injected into primary tumors and that this response was absent in doxycycline-treated tumors. The introduction of bacteria into a low-metastasis, low-microbiota mouse model (MMTV-Wnt) was sufficient to promote metastasis. Single-cell RNA-seq analysis of bacteria-invaded cancer cells showed activation of the fluid shear stress pathway. Bacteria-invaded tumor cells exposed to physiologic shear stress survived better and had less stress fiber formation than did tumor cells without bacteria. Consistent with the known role of Ras homolog family member A (RhoA) in stress fiber formation, bacterial invasion of tumor cells also suppressed RhoA and Rho-associated protein kinase (ROCK) activation. Inhibition of RhoA signaling also promoted metastatic colonization in vivo. These findings provide physiologically relevant insights into the interactions between breast tumor–specific microbiota and CTCs and how this relationship may benefit CTC survival and promote metastasis.