Abstract: Dynamic alliance and virtual enterprise (VE) are the main components of global manufacturing, of which the partner selection is a key problem in VE to the success of a project. A partner selection method or multi-objective optimization model is proposed including time, cost and risk as factors to evaluate and constraints on the project. A genetic algorithm is further developed to solve the problem. The performance of this algorithm is demonstrated by an example showing the R&D activities of a smoke detector, and the results reveal the effectiveness and efficiency of this algorithm.