Repair and rehabilitation of existing structures is a necessity compared to impractical application by direct replacement of the structures. Repair using steel plates to concrete structures had been seen as the most popular method but these techniques had drawn many problems at sites. Recent application of repair and rehabilitation involved externally bonded reinforcement that causing interface performance as an interesting area to be studied. Performance of epoxied system in externally bonded plates needs to examine more to improve database in literature. This study is required to discover the behaviour of interfacial bond in epoxied based. Bonding of external plate to concrete surfaces using adhesive is a practical solution for upgrading the existing reinforced concrete structures. This paper focuses on the bonding performance of epoxy resin adhesives to CFRP-concrete and steel-concrete prisms. Concrete prisms size of 100 × 100 × 300 mm with roughened surface on both sides was prepared. Surface preparation of the concrete prisms was carefully made prior to bonding to CFRP plates and steel plates respectively. Two groups of specimens, namely CFRP Plate-Epoxy-Concrete Prism and Steel Plate-EpoxyConcrete Prism, were prepared and tested by applying direct tensile loads until failure. The strain distribution along the bond length of the CFRP Plate-Epoxy-Concrete Prism was fairly linear but with further increase of loads, the local strain became non-uniform and non-linear. The strain distribution along the bond length for Steel Plate-Epoxy-Concrete Prism was almost linear, i.e. proportional to the applied load and uniform. The specimen did not failed at the bonding region but failed due to yielding of steel plates. Result from direct tensile test of the steel plates had drawn to conclusion that the strength of steel plates used for this study was not sufficient in term of strength. The test results showed that structural type of epoxide resin was suitable to be used as structural adhesive. The bonding strength was affected by the length of the bonding surface, modulus of elasticity of the materials and interfacial conditions. It is proven by the better performance from CFRP-prism samples with 283.2MPa local bond stress maximum differences if compared with the Steel-prism. Stress concentration was quite close at bond length of 15mm to 35mm. The stresses decreased gradually from bond length of 65mm to 155mm. The longer or wider the bonding surface, the stronger the bonding strength would be but had to be limited to effective bond length of 60 mm to 120 mm of bond length.