In this paper, a methodology for the optimal managementof a combined irrigation and water supply systembased on model predictive control (MPC) is proposed. A controlorientedmodelling methodology for this type of systems ispresented as well. MPC is used to generate flow control strategiesfrom the sources to the consumer and irrigation areas to meetfuture demands with appropriate flows, optimizing operationalgoals such as network safety volumes in dams and flow controlstability in actuators (valves, gates and pumps). The case studyof Guadiana river is used to exemplify and verify the proposedoptimal management methodology. Results have shown the effectivenessof the proposed modelling and control methodologies.