The state-space explosion problem, resulting from the reachability computations in controller synthesis, is one of the main obstacles preventing supervisory control theory from having an industrial breakthrough. To alleviate this problem, a strategy is to symbolically perform the synthesis procedure using binary decision diagrams. Based on this principle, the work presented in this brief develops an efficient symbolic reachability approach for discrete event systems that are modeled as finite automata with variables, referred to as extended finite automata. Using a disjunctive event partitioning technique, the proposed approach first partitions the transition relation of the considered system into a set of partial transition relations. These partial transition relations are then selected systematically to perform the reachability analysis, which is the most fundamental challenge for synthesizing supervisors. It has been shown through solving a set of benchmark supervisory control problems for EFA that the proposed approach significantly improves scalability in comparison with the previously published results.