Brain metastases occur in 20 to 40% of patients with metastatic breast cancer. The process is complex and depends on successful cancer cell evasion from the primary tumor, distribution and survival within the blood stream and cerebral microvasculature, penetration of the blood brain barrier and proliferation within the brain microenvironment. The initial steps of brain colonization are difficult to study in vivo. Therefore, in vitro assays have been developed to mimic this process. Most commonly, in vitro studies of brain colonization focus on tumor cell adhesion to brain endothelial cells and transendothelial migration. We previously investigated breast cancer brain colonization from the blood stream in vivo and defined the time and process of brain entry for five different cancer cell lines in a mouse model. We now investigated if in vitro approaches can reliably emulate the initial steps that determine successful brain colonization in vivo. To this end, we optimized an in vitro model of the vascular blood brain barrier and compared the brain invasion properties of the in vivo characterized cell models with their ability to interact with and penetrate the blood brain barrier model in vitro. Our results show that the in vitro findings correlate only poorly with the vivo results. The limitations of the in vitro approaches are discussed in light of the in vivo processes. We conclude that investigation of mechanisms supporting the earliest steps of breast cancer brain metastasis from the blood stream will depend on in vivo analyses.