The current approaches authorized by Eurocode 2 for the design of reinforced concrete (RC) structures at the structural scale are: linear-elastic analysis, linear-elastic with limited redistribution of moments, plastic and non-linear analysis. Numerous experiments on RC beams were performed which provided interesting information for the models proposed in the standards. However, at the RC frame level where the beam-column are present, although several experiments have already been carried out in the literature, this topic is still to be investigated, due the important number of parameters: interior joint, exterior joint, stiffness ratio between beam and column, steel reinforcement type, … This paper presents firstly an experimental study on an H-form RC frame structure (two vertical columns and one horizontal beam). Several loading-unloading cycles were carried out. Displacement fields during the test were measured by the image correlation technique. Then, experimental results were compared with that of the analytical models proposed in Eurocode 2 and an advance FE code (CAST3M). The comparisons show that analytical methods presented in Eurocode 2 underestimated about 20–30% of the ultimate capacity of the structure; while the non-linear analysis with CASTEM code – when material characteristics and boundary conditions are correctly identified - could provide good results comparing to the experiments.