Henan Province is located at the Central Plains of China. It is not only a transition zone from subtropical to warm regions, but also a transition zone from east plain to west mountanous area. Hence it has a complex ecosystem with high species diversity. Through more than ten years′ investigation, a total of 8637 insect, spider and mite species in Henan Province have been recognized. These species belong to 3967 genera of 551 families in 32 orders, accounting for 11.2% of the total species in China and 0.8% in the world. By using the multivariate similarity coefficient formula and the multivariate similarity clustering analysis (MSCA) proposed by the present authors, all the known species of insects, spides and mites along with a great number of data of the distribution sites were analyzed in Henan Province. The general similarity coefficient of 11 ecotypes in the whole province was 0.184, showing an asymmetric distribution. At the similarity level 0.3, the whole province was divided into four geographic regions, which were Taihang mountainous area in the northwest and the hilly platform area in the west, the north and south slopes of Funiu Mountains, the Tongbo Mountainous area and its adjacent Dabie mountainous area, and the plain and basin region. Clustering the geographically close and topographically similar areas into different distribution regions was very natural and rational, which showed that the analysis method was scientific and the geographical unit and distribution data were applicable. It is superior to the traditional method of clustering analysis, which requires gradual combination and degradation. It should be noticed that, as an accepted boundary between the palearctic and the oriental regions, the Funiu Mountain-Huai River did not give play to the geographical division of Henan as expected. The fact that the south and north slopes of Funiu Mountain was steadily clustered into one region and both banks of Hui River was not divided into two regions showed at least that mountains with an altitude of 2000 m were unable to become distributional barriers of insects, and the narrow Huai River was unlikely to cause biological isolation. To explore the factors affecting the integral similarity, the influence of different taxonomic categories and groups, different fauna elements and scales of distribution sites on the similarity clustering relationships of the whole province was analyzied. The result showed that the increase of the taxonomic categories could increase the similarity and conseal differences; the category at the genus level could effectively reveal the similarity relationships of different ecotypes, while at family level it was not appropriate in analyzing the inter-provincial geographic regions, but could adapt to larger geographic regions. Each fauna has its own distribution patterns, but all except widely spreding species can reveal the similarity clustering relationships that are not very different from the general characters. The biological group is the largest parameter influencing the similarity clustering structure. It is the key factor to open out the macrocosm law by ensuring more groups taking part in. The numbers of species and the distribution sites are not the sensitive factor under the premise that more groups can participant in. Even 10% species can fully post the general clustering characters as long as the increse or decrease of certain species with definite distribution patterns are not involved. So in the similar research, intentionally eliminating the species with definite distribution patterns is unadvisable. Distribution sites are the basic material in calculating similarity. When enough categories and number of species participate in, the slight increase or decrease in the number of distribution sites can only result in a moderate increase or decrease of comparability coefficient, usually will not result in the change of clustering structures. It can be assumed that the basic investigation of the insects fauna and their distribution on both national and provincinal scales has provided enough data to support the similarity analysis and the geographic division. When most groups participate in, it is not necessary to rigidly adhere to include all the species and all the distribution sites, not to say the undescribed species and the uninvestigated sites. The capacity to treat a large quantity of data and the convenience of MSCA method used in this study can provide sufficient technical support to the biogeographic study.