SUMMARYSimilar to circulating tumour and immune cells, many blood-borne microbes preferentially “home” to specific vascular sites and tissues during hematogenous dissemination 1–5. For many pathogens, the “postal codes” and mechanisms responsible for tissue-specific vascular tropism are unknown and have been challenging to unravel. Members of the Lyme disease Borreliella burgdorferi species complex infect a broad range of mammalian tissues and exhibit complex strain-, species- and host-specific tissue tropism patterns. Intravenous perfusion experiments and intravital microscopy studies suggest that heterogeneous tissue tropism properties may depend on tissue-specific differences in host and microbial molecules supporting vascular interaction and extravasation. However, interpreting these studies can be complicated because of the immune-protective moonlighting (multitasking) properties of many B. burgdorferi adhesins. Here, we investigated whether B. burgdorferi vascular interaction properties measured by live cell imaging and particle tracking in aorta, bladder, brain, joint and skin microvascular flow chamber models predict strain- and tissue-specific dissemination patterns in vivo These studies identified strain- and endothelial cell type-specific interaction properties that accurately predicted in vivo dissemination of B. burgdorferi to bladder, brain, joint and skin but not aorta, and indicated that dissemination mechanisms in all of these tissues are distinct. Thus, the ability to interact with vascular surfaces under physiological shear stress is a key determinant of tissue-specific tropism for Lyme disease bacteria. The methods and model systems reported here will be invaluable for identifying and characterizing the diverse, largely undefined molecules and mechanisms supporting dissemination of Lyme disease bacteria. These methods and models may be useful for studying tissue tropism and vascular dissemination mechanisms of other blood-borne microbes.