A trend towards a more relational type of contracting can be noticed in the Dutch infrastructure industry. These so-called relational project delivery models are supposed to deal with the current challenges of regular cost and time overruns, the growing demand and the adverse relationship between contractor and client in traditional projects, leading to an unfair risk allocation, lack of collaboration and lots of conflicts. One of these measures proposed by RWS is the introduction of the two-phase model. For specifically the first phase it is unclear how the collaboration can be improved and how a collaborative identity might contribute. The research question therefore is: How could the first phase of the two-phase model in infrastructure projects in the Netherlands be organized to improve collaboration and how does the formation of a collaborative identity contribute to this? A further operationalization of collaborative identity resulted in a definition: An identity within the project that distinguishes the organization and contributes to a collaborative mindset in the project organization. Improved collaboration can be achieved by implementing certain principles in the project by the organization of specific working practices. These working practices and principles were gathered from different streams of literature such as the alliance, early contractor involvement and bouwteam models. The principles are scored on their importance to collaboration in the first phase of the two phase model and the working practices were scored on their contribution to achieving the principles. This resulted in a framework identifying the most important principles and working practices. This framework was then discussed with two separate focus groups from the contractor and the client. The comments and discussion was used to update and improve the framework. The framework is found in figure A. The most important conclusions were to apply the framework in a so called project start up. Furthermore a shared vision and objectives related to the project are important for the two-phase model, as two organizations should collaborate intensively in the first phase, while the price is not yet set. Involving the right people who can work in this ‘new’ way in the project is vital, and the price should be constructed using open book, and avoiding fixed price elements in the tender phase. Finally it is concluded that the collaboration in any project should be an active project goal. This means that effort should be made to evaluate, discuss and reward behaviour resulting in good collaboration. The conclusions indicate the importance of a collaborative identity to the collaboration in the first phase of the two phase model and this was also specifically confirmed by the participants of the focus group. For future research it is recommended to perform more research on collaborative identity. Secondly more research should be performed on the two-phase model, when the projects have been finished, also relating to the success of the overall project. Finally research on the collaboration in the second phase and the transition from the first to the second phase should be performed.