A module partition approach based on density algorithm and design structure matrix is proposed for solving the problem of product modular design, considering the diversity of design structure matrix in actual pro- cess. Design structure matrix is utilized to describe the relationship between components, which are clustered according to the density reachable principle, by judging the Euclidean distance of the n-dimensional space. Dif- ferent partitioning results are obtained while adopting the different algorithm parameters, and the modularity metric is acted as the evaluation index of the clustering results to select the optimal module partition scheme. Finally, the module division process is illustrated by taking the truck crane as an example.