Among all alterations that cells need for colonizing distant organs, migratory proficiency is widely accepted as a key characteristic of highly metastatic cells. However, investigations are almost exclusively based on functional inhibition or genetic knock-out of candidate proteins involved in cell migration to study their impact on in vivo metastasis. Importantly, there is no antibody or gene that allows the labeling of highly migratory cells. Identification and capture of live individual cells in a heterogenous ensemble typically rely on genetic manipulations that encode fluorescent probes. Thus, a precise understanding of how several molecular components interact to yield the phenotype of interest is a prerequisite to distinguish and isolate such target cells. Single cell magneto-optical capture (scMOCa) has the capacity to tag and manipulate cells based on visual phenotypes to subsequently identify and characterize novel molecules that modulate these phenotypes. Here, we capture single cells from a highly heterogeneous triple-negative breast cancer cell population based solely on the analysis of their migration trajectories. We isolate and expand intact cells to generate novel populations defined by their travelled distance alone. We demonstrate that the progeny of these cells stably retains their migratory phenotype and exhibits a dramatically different transcriptomic profile when compared to the parental counterpart. Comparison of RNA sequencing of fast and parental cell populations reveals a motility-related transcriptomic profile that highlights Biological Adhesion Gene Ontology terms. Fast cells possess an increased dynamics and distinct morphology of focal adhesions. Finally, primary tumors established from fast cells generate a higher number of circulating tumor cells and soft tissue metastases in vivo, resulting in shorter time to clinical endpoint following tumor resection. Thus, selection for a highly migratory phenotype alone can confer increased fitness for specific steps in the metastatic cascade.