Abstract: Technique for order preference by similarity to ideal solution (TOPSIS) was improved, through replacing euclidean distance with vertical distance and adding weight to vertical distance. Taking the choice of electric insulating materials used in aerospace engineering as an example, 5 kinds of candidate materials were comprehensively evaluated by improved TOPSIS according to 8 kinds of assessment indexes selected from the point of functionality. The results show that the weighted vertical distance of quartz glass was the shortest, which meant it was the best candidate material, and the evaluate results were correct. Improved TOPSIS resolved the shortcoming of the conventional TOPSIS, showed the different situation of candidate materials better, and displayed a clear physical implication, with which the analysis was more comprehensive and objective. Hence, improved TOPSIS was a tool for making decision in engineering material selection.