Computer aided history matching techniques are increasingly playing a role in reservoir characterization. This article describes the implementation of a differential evolution optimization algorithm to carry out reservoir characterization by conditioning the reservoir simulation model to production data (history matching). We enhanced the differential evolution algorithm and developed the modified differential evolution optimization method with random localization. The proposed technique is simple-structured, robust, and computationally efficient. We also investigated the convergence characteristics of the algorithm in some synthetic oil reservoirs. In addition, the proposed method is compared with the Nelder-Mead simplex search method and a standard Genetic algorithm.