Matrix metalloproteinases (MMPs) are a family of endopeptidases that degrade extracellular matrix (ECM) proteins, functioning in various physiological processes such as tissue remodeling, embryogenesis, and morphogenesis. Dysregulation of these enzymes is linked to multiple diseases. Specific inhibition of particular MMPs is crucial for anti-MMP drug development as some MMPs have shown anti-disease properties. In this study, we aimed to design a highly specific inhibitor of MMP-9, that plays a crucial role in cell invasion and metastasis, using tissue inhibitor of metalloproteinases 2 (TIMP2), an endogenous broad-family MMP inhibitor, as a prototype. In our earlier work, we were able to narrow down the specificity of the N-terminal domain of TIMP2 (N-TIMP2) toward MMP-9, yet at the expense of lowering its affinity to MMP-9. In this study, a library of N-TIMP2 mutants based on previous design with randomized additional positions was sorted for binding to MMP-9 using yeast surface display. Two selected N-TIMP2 mutants were expressed, purified and their inhibitory activity against a panel of MMPs was measured. The best engineered N-TIMP2 mutant (REY) exhibited a 2-fold higher affinity to MMP-9 compared to that of the WT N-TIMP2, and 6- to 1.1x104-fold increase in binding specificity toward MMP-9 compared to five alternative MMPs. Moreover, REY demonstrated a significant increase in inhibition of cell invasion and proliferation compared to the WT N-TIMP2 in MDA-MB-231 breast cancer cells. Therefore, our engineered N-TIMP2 mutant emerges as a promising candidate for future therapeutic development, offering precise targeting of MMP-9 in MMP-9-driven diseases.