Abstract: To match the three dimension digital model and the CAD model, a new method is proposed, which consists of the surrounding box -based initial registration and the float data coded adaptive genetic algorithm-based accuracy registration. The initial registration can ensure the ranges of the spatial transform parameters for the accuracy registration. In the accuracy registration, the goal function is constructed by the method of the least squares. In order to overcome the premature convergence, an index that is related with the diversity of the population is defined. The index is used to define the crossover operator and mutation operator in the genetic algorithm, and to adjust the probabilities of the crossover and mutation adaptively. The result showed that the method can properly match the three dimension digital model and the CAD model, and the algorithm is steady and reliable.