The PhD research project “MultiCoSteel” is focused on a validation of the Vierendeel theory and a development of an innovative, easy-applicable and efficient type of shear connector for an application in composite columns with multiple-encased steel profiles. A complex interaction between the materials and lack of knowledge regarding composite behaviour opens opportunities to develop, improve and optimise the system and the design method. Within this research project, both aspects are handled within the objectives; 1) to develop a new type of shear connector and 2) to prove the design hypothesis of Vierendeel mechanism behaviour.Composite structures are the most efficient building systems nowadays. It has to be however ensured that all the elements work together by transferring forces, especially shear stresses. Till now, the composite action was mainly ensured by applying shear connectors in form of the shear studs. This type of shear connection was developed for composite beams, therefore there is still big potential to develop more efficient type of shear connector for composite columns. A new type of the shear connector has a form of modifications applied on the flanges of the steel profiles. A preliminary numerical study was carried out to identify the most performing solution. In respect to the mechanism and behaviour of the large composite columns, the common practice so far is to assume homogenous system with one stiffness and analyse it based on the Bernoulli beam theory. However, the results of the recent tests caused some doubts whether this mechanism is correct. Strong analogy was observed to the system described by the Vierendeel truss theory. It is not clear which of the mechanisms describes better the behaviour of composite columns and what the limitations of using one or the other are. This research project aims to identify the most suitable model and design approach.