The hydrothermal coordination can be defined as a problem to determine the optimum usage of the hydroelectric and thermoelectric resources available during a period. In hydrothermal generation systems with a predominance of hydroelectric power plants, like in the Brazilian system, the problem consists in replacing the thermal generation by hydropower generation to minimize the system operational costs. Therefore, this paper presents a model based on Particle Swarm Optimization, and Network Flow applied to the problem of hydrothermal coordination. The goal is to determine an optimal operational strategy for the reservoirs of the hydroelectric power plants considering each hydroelectric power plant separately, its operational constraints and guaranteeing the applicability of the solutions, in order to minimize the operating cost of the system. The proposed approach is compared with four other optimization methods: a Genetic Algorithm (GA), a model based on Genetic Algorithms and the Takagi-Sugeno Fuzzy Inference System (GA+Fuzzy), a model based only on PSO and an optimization algorithm that employs Network Flow and Reduced Gradient (NF+RG). A hydroelectric system composed of three hydroelectric power plants and six different hydrological scenarios were used to test the algorithms. The primary objective was to illustrate the viability and applicability of the proposed algorithm, and, based on the obtained results, show the efficacy and energy gains that are possible.