Inhibition of angiogenesis is an effective and low toxic therapeutic avenue for the treatment of cancer patients in addition to traditional interventions. Majority of current available angiogenesis inhibitors for cancer therapies are growth factor inhibitors and small molecule tyrosine kinase inhibitors. A number of endogenous proteins and/or proteolytic fragments of extracellular matrix proteins are shown to have the activity of inhibition of angiogenesis by directly targeting endothelial cells. Structural analyses have indicated that a common structure of anti-parallel β-sheet with a highly positively charged surface presents in many of those inhibitors. This common structural feature is critical for the maintenance of their anti-angiogenic function. With this structural information, we have designed and developed a new class of anti-angiogenic proteins by integrating the short anti-parallel β-sheet forming sequences of endogenous anti-angiogenic proteins into a stable host protein, the extracellular domain-1 of cluster of differentiation 2 molecule (CD2D1). 1D 1H NMR spectra analyses indicated that the designed anti-angiogenic protein (ref to as ProAgio) folded as a β-sheet structure similar to that of the parental protein, CD2D1. ProAgio inhibited the growth of human umbilical vein cells (HUVECs) without affecting the growth of epithelial cells, suggesting a specific effect to endothelial cells. ProAgio effectively reduced endothelial tubules formed by the co-culture of HUVECs and PC3 cells on matrix gel in vitro. The designed anti-angiogenic protein was further site-specifically PEGylated in order to improve PK/PD properties and reduce immunogenicity. Examinations with PC3 xenografts showed that both ProAgio and the PEGylated ProAgio dramatically inhibited tumor growth. Immunofluorescence staining analyses of the endothelial marker CD31 indicated dramatic decreases in tumor vessels in lengths and branching points. Histological and immunofluorescence staining analyses of tissue slices of major organs indicated that there were no pathological damages to the tissue structure or disruption of normal vessels associated with the treatment of our designed anti-angiogenic agent. Overall, our studies developed a novel anti-angiogenesis agent that may have great clinical potentials. Our concept of protein design can be extended to the development of other novel protein drugs.