This article describes a method for optimizing injection rates in fractured geothermal reservoirs. The optimization approach suggested here is based on maximizing the net present value (NPV) of production revenue from the reservoir. The method relies on tracer and flow-rate data, which are used to calibrate a thermal transport model in such a way that the thermal drawdown depends on the injection rates. Then, an empirical correlation is used to relate the injection and production temperature to the specific electrical power output. The power output model, in conjunction with predictions for the future energy prices and interest rates, allows the computation of the NPV as a function of the injection rates. This characterization of the problem makes it possible to solve it quickly and efficiently. The optimization method was applied to two discrete-fracture reservoir simulation models. The first having a relatively simple structure with two injectors and two producers and the second being a more complex model with seven wells and a fracture structure based on observations from the Soultz-sous-Forets enhanced geothermal system (EGS) in France. The results show that the method works well for both the simple and more complex case. A finding of particular interest is that the optimal injection schedule depends strongly on the minimum design temperature for the power plant. (C) 2013 Elsevier Ltd. All rights reserved.