This thesis presents the behavior of a partial strength joint between a tubular column and an I-section beam under monotonic and cyclic loads. Extended endplate and blind bolts are used to form the connection between the beam and column. The use of tubular-section profiles is becoming noteworthy nowadays in steel frames because of their greater torsional and flexural performance than open-section profiles. However, the design of tubular-section column to beam connection is not sufficiently covered in the current design specifications, such as EC-3 and EC-8. The component-based design procedure provided in EC-3 is only applicable for joints made to open-section columns. The thesis aims to explore the potential behavior of the connection under monotonic and cyclic loads using a non-linear 3D finite element analysis. A total of twenty-five cases are developed to investigate the effects of different components of the joint on the behavior of the beam-to-column connection system. Validation of the finite element is first carried out for both load cases, and then the differences of the welded and extended endplate connections are examined prior to the detailed parametric studies. The influences of three stiffening methods (i.e., inner diaphragm, infilled-concrete, and thickening column wall), axial column load, endplate thickness, and bolt size are investigated. Detail measures of the failure mode, strength, initial stiffness, joint classification, ductility, stiffness and strength degradations, and energy dissipation capacity of the joint were taken to characterize the joint behavior. The results of welded and extended-endplate connections that have approximately equal maximum moment capacity, 14.4% higher initial stiffness and 14.24% lower rotation capacity are observed in the extended-endplate connection during monotonic load. However the differences observed in the initial stiffness and rotation capacity for the two connection types are negligible during cyclic load. The total energy dissipated in the welded joint is normally larger than endplate joint. The inner diaphragm, thickened column wall and concrete-filler have enhanced the initial stiffness by 68.32%-84.36%, 57.3%-120% and 430%-530%, respectively. The energy dissipation capacity of the joint has also showing a positive improvement in the stiffened joints. Unlike the first two stiffening methods, the concrete-filler have also improved the ductility of the joint by about 28.4%-52.4%. vii Increasing endplate thickness has enhanced the rotational stiffness of the joint by 47.6%133.6%, but its effect in terms of improving the strength of the joint has only showing 2.8%8.1% and 20.8%-26.1% progress for the monotonic and cyclic loads, respectively. The change in endplate thickness has also influenced the ductility of the joint. Differently, it is observed that the strength of the joint have been increased about 20.3%-58.4%, by increasing the size of the bolt. The rotation and energy dissipation capacity of the joint have also displayed progress as the bolt size increased. Failures that governed by column plastic deformation are generally produced higher ductility and lower energy dissipation capacity in comparison to failures governed by beam inelastic buckling. From the variation of column axial load, it is observed that the joint is not sensitive to different levels of axial loads in the column.