As a new Web pattern, Web service has been rapidly developed in recent years. How to dynamically integrate the existent Web services to form a newly value-added and complex service to meet the requirement of different users is a popular research area. This paper presents an algorithm GODSS (global optimal of dynamic Web services selection) to resolve dynamic Web services selection with QoS global optimal in Web services composition. The essence of the algorithm is that the problem of dynamic Web Service selection with QoS global optimal is transformed into a multi-objective services composition optimization with QoS constraints. The theory of intelligent optimization of multi-objective genetic algorithm is utilized to produce a set of optimal Pareto services composition process with constraint principle by means of optimizing various objective functions simultaneously. Theoretical analysis and experimental results indicate the feasibility and efficiency of this algorithm.