Background: Breast cancer (BC) is highly prevalent in women and is often fatal. Vasculogenic mimicry (VM) is associated with cancer progression. The expression of Cavolin-2 (CAV-2) affects the occurrence and development of BC, but the specific mechanism of CAV-2 in BC is still poorly understood. The aim of our study was to explore the specific mechanism of the effect of CAV-2 on the occurrence of VM and the malignant biological behavior of BC. Methods: First, we detected the gene expression of CAV-2 in BC tissues and cell lines. The overexpressed CAV-2 plasmid (oe-CAV-2) and si-CAV-2 were then transfected into michigan cancer foundation-7 (MCF-7) cells to determine the effects of different expression levels of CAV-2 on BC cells. In addition, to further determine the feedback loop mechanism between CAV-2 and epidermal growth factor receptor (EGFR), the cells were treated with (1) AG1478, which inhibits EGFR phosphorylation, (2) an EGFR phosphorylation activator epidermal growth factor (EGF), (3) a ubiquitination proteasome inhibitor MG-132, and (4) an extracellular regulated protein kinases1/2 (ERK1/2) inhibitor SHR2415. In the cell experiment, VM and the biological behavior of BC cells were mainly observed by tumor spheroid formation, tube formation, Transwell, 5-ethynyl-2-deoxyuridine (EdU), and scratch tests. In addition, we performed Western blot analysis of epithelial–mesenchymal transition (EMT)-associated proteins and angiogenesis analog markers. Finally, subcutaneous injection of breast cancer cells into nude mice was used to establish an animal model of BC so that the effect of CAV-2 on the progression of BC could be verified in vivo. Results: CAV-2 was reduced in BC patient cancer tissue and in human BC cells (p < 0.001). Overexpression of CAV-2 effectively inhibited the progression of BC cells and reduced the formation of tumor spheres and tubules, whereas knockdown of CAV-2 had the opposite effect. CAV-2 and EGFR could interact, and the expression of EGFR was higher after overexpression of CAV-2 (p < 0.001), but P-EGFR expression was lower (p < 0.01). Treatment with the EGFR phosphorylation inhibitor AG1478 further enhanced the effect of oe-CAV-2. Treatment with the EGFR phosphorylation activator EGF reversed the effects of oe-CAV-2. After EGF treatment, the CAV-2 ubiquitination level was increased, but its gene expression decreased (p < 0.01). EGF promoted VM occurrence and malignant biological behavior of BC cells, and the EGF effect was reversed after overexpression of CAV-2. EGF also promoted the upregulation of P-ERK1/2 expression (p < 0.001), and the promotion effect of EGF on the behavior and angiogenic mimicry of BC cells was reversed after treatment with SHR2415, an ERK1/2 inhibitor. Animal experiments have also confirmed that CAV-2 can inhibit tumor growth and the phosphorylation of EGFR and ERK1/2, and this effect can also be reversed by treatment with EGF to a certain extent. Conclusion: CAV-2 can inhibit the phosphorylation of ERK1/2 by antagonizing EGFR phosphorylation and downregulating the expression of p-EGFR, thereby inhibiting VM and the malignant biological behavior of BC. In the process of regulating EGFR phosphorylation by CAV-2, p-EGFR can induce CAV-2 ubiquitination, that is, CAV-2 regulates the ERK signaling pathway through an EGFR feedback loop and influences the angiogenic mimicry and malignant biological behavior of BC.