The paper presents a research 1that aims to develop a methodology that supports the elaboration and the optimisation of long term maintenance and renewal policies for railway infrastructure (strategic planning level, up to 20 years). The elaboration process takes the capacity, in other words the costs of track possessions, into account. The developed methodology is divided into two steps: the optimal track possession estimation and the maintenance and renewal policy elaboration. A long term maintenance and renewal policy sets the decision framework that applies to the medium-term planning process, which defines the maintenance and renewal program (tactic level, up to 6 years). This decision framework avoids that planning works only focus on a medium-term optimisation, without considering the life-cycle and the long term evolution of the infrastructure (and thus leading to a fatal loss of substance). The decision framework sets the part of infrastructure that should be yearly maintained and renewed, at the corridor or network level, as well as the type of material to use at the time of renewing components. The first step provides the track possession strategy that minimise the costs of the maintenance and renewal actions. The model receives as input the costs of the track possession, calculated with a capacity assignment model (see : "Capacity Evolution Modelling for Long Term Planning", Moreira N., et al), and the structure of the network. The track possession strategy defines then the costs of maintenance and renewal actions, including the track possession ones. On the second stage, the long term maintenance and renewal policy is evaluated by applying the corresponding decision framework to a simulation of maintenance and renewal needs engendered by the traffic scenarios. The paper includes some results from a case study where the model has been applied to a strategic railway corridor, in Switzerland, showing the impact of the cost of capacity (track possessions) over the long term maintenance and renewal policy.