Matrix metalloproteinases (MMPs) are considered excellent targets for cancer therapy because of their important roles in multiple aspects of tumor growth and metastatic spread. However, not all MMPs, or even all activities of specific MMPs, promote cancer. Therefore, there is a need for highly specific inhibitors. Monoclonal antibodies provide the potential for the degree of specificity required, but the isolation of antibodies able to inhibit a specific protease with high selectivity is challenging. Proteolysis specificity lies in recognition of the substrate in or around the active site, which generally forms a concave cleft inaccessible by human IgGs. Inspired by camelid antibodies, which have convex paratopes, we have produced a recombinant human IgG, designated 3A2, which binds in the substrate cleft of MMP-14, inhibiting its activity, but not the activity of highly homologous MMPs. In the 4T1 highly metastatic, syngeneic, orthotopic model of breast cancer, IgG 3A2 markedly inhibited growth of the primary tumor, but more importantly reduced metastatic spread to the lungs and liver by 94%. Stem cells in the tumor population expressed twice as much MMP-14 mRNA as bulk tumor cells. In addition to reducing dissemination of tumor stem cells, as would be expected from inhibition of MMP-14's ability to degrade components of the extracellular matrix, IgG 3A2 also inhibited the ability of individual stem cells to proliferate and produce colonies. We conclude that it is possible to produce antibodies with sufficient specificity for development as therapeutics and that IgG 3A2 has therapeutic potential.