In the modern manufacturing and operations management, on-time delivery is a critical factor towards realizing customer satisfaction. This paper focuses on job-shop scheduling problem to minimize total weighted tardiness and proposes a discrete differential evolution algorithm for this problem. In order to improve the search ability and efficiency, this paper hybrids the local search which is based on the scheduling critical path theory, and develops a migration operation based on the population diversity theory to jump out of local optimum. Computational results on benchmark instances from the literatures show that the proposed algorithm can compete with the existing algorithms.