Congestion is a major limiting factor preventing expansion of renewable energy production in distribution networks. However, with large shares of connected power electronic-interfaced generators in combination with new types of controllable loads, such as electric vehicles (EVs), there is a potential to greatly increase network operation flexibility. Utilising these available flexible resources effectively is crucial to boost network capacity in a cost-effective manner and allow for safe integration of additional renewable energy sources (RESs). In parallel, the reactive power flows in distribution networks are changing. This can be attributed to the increased RES production and increased charging currents due to expanding cable networks. Also contributing to the changing flows is the rising number of new household appliances and consumer electronics with non-linear load characteristics. This makes systemwide coordination of resources an even more pressing issue. For distribution system operators (DSOs), minimising undesired reactive power flows at the connection to the transmission system is key to meet inter-network requirements. In this paper we propose a centralised near real-time control algorithm for combined congestion management and reactive power control in distribution networks. Through updated communication and measurement protocols, together with more extensive use of the active and reactive power control capabilities of local flexibility resources – such as wind power plants (WPPs), photovoltaic (PV) units, and flexible loads – bottlenecks can be detected and eliminated. Flexibility is offered by local resources and dispatched by the DSO through a common platform, which is independent of any specific financial arrangement for the participating flexibility providers. Thus, market solutions and individual contractual agreements are not mutually exclusive and can be implemented in parallel. The inclusion of reactive power simplifies the DSO’s coordination of intra-network and inter-network operational requirements. We demonstrate selected algorithm features through simulations of a congestion scenario in a medium voltage benchmark network. Aspects of deploying the solution in actual distribution network operation are also outlined.