Abstract: Automated generalization in condition of digital mapping is an international problem, conventional model and algorithms can not fulfill its requirement now. On the base of that most problems of cartographic generalization can be treated as optimizing problem, this paper carries genetic algorithm into the study of cartographic generalization in a creative way and builds a model of cartographic generalization based on genetic algorithm. Taking the selection of point as example, this article lucubrates the keys of the model—coding, fitness function, etc. Examples with actual experiment demonstrate that the model of cartographic generalization based on genetic algorithm is a new and feasible method to solve the problem of automated generalization in condition of digital mapping.