Due to the lack of capacity design principles as well as of appropriate structural details, most of the reinforced concrete building designed primarily for gravity loads as typical of pre- 1970s code provisions, are expected and has been demonstrate to suffer sever damage or total collapse under the earthquake excitation. Due to the use of plain round bar and inadequate reinforcing details, critical shear failure in the joint connection region could occur, leading to sever damage when not total collapse of the building. In this research project, a comprehensive experimental programme was carried to investigate the seismic performance of existing beam column joints prior and after retrofit intervention with a recently proposed low-invasive retrofit technique based on a metallic haunch system. The joint performance was evaluated in terms of the principal tensile stresses that caused the joint shear cracks in the joint panel zone. Quasi-static cyclic tests under uni-directional or bidirection loading regime were carried out to record the response of a series of under-designed beam column joints (with either a wide-beam or a deep-beam solution, deformed or plain round bars with end hooks). The experimental results were used to investigate the effect of structural detailing and loading regime on the seismic performance. To retrofit the potential deficiencies in the existing beam-column joints, the feasibility and efficiency of a low invasive retrofit solution based on a diagonal metallic haunch was investigated. The proposed haunch retrofit solution aims to provides an economic, ease of implementation alternative to protect the joint from the brittle shear failure by relocating the beam plastic hinge away form the joint panel zone. To achieve the desired capacity design (hierarchy of strength) and sequence of event, a simplified analytical formulation has been adopted to account for the joint shear strength in terms of principle tensile/compression stresses prior and after the retrofit intervention. A useful visualization tool based on a M-N (moment-axial load) performance domain can be adopted to evaluate the actual performance point and events, by comparing demand vs. capacity. Designed charts are proposed based on displacement compatibility conditions to evaluate the efficiency of the haunch solution. In addition, a complete step-by step design procedure to implement the retrofit strategy and intervention to achieve the desired hierarchy of strength, by using the proposed diagonal metallic haunch solution, is derived and presented. The effectiveness of the proposed haunch solution and reliability of the derived analytical design/assessment procedure, were validated through experimental tests of 2-D and 3-D subassemblies, shown in the first experimental part to have the most vulnerable behaviour in the joint panel zone. Conceptual issues related to the design of the retrofit intervention, when moving from a 2-D to a 3-D behaviour are discussed. The experimental results showed an excellent performance of the proposed intervention, able to protect the panel zone region (by limiting the principle tensile stress demand), while enforcing the formation of a plastic hinge in the beam, far away from the joint interface. As a result, a much more stable inelastic response could be developed, confirming the high potential of such a low-invasive, low-cost retrofit intervention on under-designed frame systems. In conclusion, a simple numerical model, based on a lumped plasticity approach, was developed and validated on the experimental results to capture the full response of the subassembly prior and after the retrofit intervention.