User�s sensitive image to products is the emphasis that industrial design concerns, and is also the one of the key targets. Complicated mapping relations exist between design elements and user image, which make product�s user image unpredictable. The paper tries to build a model to describe the mapping relations on the basis of graph theory, which introduces the concept of genetic regulatory network (GRN) into industrial design. The model takes design elements as nodes and the relationship between elements as edge to build and draw product�s GRN graph. Through the GRN, product�s design elements are classified and the types of nodes and nodes groups are recognized, as abstracted knowledge for designers to arrange their design activities, or prepare the searching strategy for automatic optimizing programs. An example of water bottle design is carried out to test GRN�s aiding effect to product design.