DC power flow is a commonly used tool for contingency analysis. Recently, due to its simplicity and robustness, it also becomes increasingly used for the real-time dispatch and techno-economic analysis of power systems. It is a simplification of a full power flow looking only at active power. Aspects such as voltage support and reactive power management are possible to analyse. However, such simplifications cannot always be justified and sometimes lead to unrealistic results. Especially the implementation of power flow controlling devices is not trivial since standard DC power flow fundamentally neglects their effects. Until recently, this was not an issue as the application of power flow controlling devices in the European grid was limited. However, with the liberalisation of European electricity market and the introduction of large wind energy systems, the need for real power flow control has emerged and therefore, the use of these devices has been reconsidered. Several phase shifting transformers (PST) are being installed or planned in order to control flows. Therefore, it is important to fundamentally re-validate the fast, but less accurate, DC power flow method. In this paper the assumptions of DC power flow are analysed, and its validity is assessed by comparing the results of power flow simulations using both the DC and AC approaches on a modified IEEE 300 bus system with PSTs.@en