We present an improved version of Honey Bees Mating Optimization (HBMO) algorithm to develop operating rules for multi-reservoir systems. The performance of the proposed model is tested through sensitivity analysis and comparing the result with those of a real-coded Genetic Algorithm (GA) for a 60-month period single-reservoir operation problem. The improved model is subsequently employed to derive release rule and storage balancing functions which form operating policy for a multi-reservoir system along two case examples: (i) water supply and (ii) hydropower generation. The obtained operating rule curves can be used to guide the system operators in decision-making. These rule curves provide the operator with the opportunity to systematically look at the system and to make proper decisions. The obtained results showed that the optimization technique proposed in this study is capable of solving complex multi-reservoir systems operation problems. Moreover, the proposed structure properly handled the tight constraints defining the parallel reservoirs operation in such a way that all the generated solutions were feasible after a particular set of iterations. The proposed optimization algorithm of this study can be developed more in future to solve multi-modal optimization problems, and also to define operation policies for highly complex multi-reservoir systems.