Abstract: Internet as a new kind of geographical factor represents an exclusive regional differential features, it attracts more and more concerns of geographers and other discipline researchers as well. Internet in China shows an overwhelming increasing trend recently, though this momentum was commenced late in 1994. According to the report of the China Internet Network Information Center (CNNIC), by December 31, 2002, in Mainland China, there are 59.1 million Internet users, 20.83 million computers connected to the Internet, 179.544 thousand domains registered under ‘CN’, and 371.6 thousand websites under ‘www’. As in many western countries, Internet in China has unevenly diffused from the eastern to the western region and among 31 mainland provincial units. Based on the regularly released data by CNNIC, this study explores regional differential of Internet diffusion in China by quantitative methods. The authors design an Internet Diffusion Index (IDI) to stratify Internet diffusion level of the 31 provincial units. IDI is the geometric mean of 3 indicators; they are Internet users per thousand people, the domain names per 10 thousand people and the websites per 10 thousand people. Given the national average IDI is 1.000, the final results show that the IDI of the eastern coastal region is 1.754, it is much higher than 0.389 of the central region and 0.422 of the western region. However it is easy to find that IDI of the west is slightly higher than that of the central region. The IDI topology is different from that of social and economic development, which has new implication for China's regional development. As to the 31 provincial unites, 6 groups are further sorted out, Beijing and Shanghai rank the highest group, only Guizhou is in the lowest group, 6 coastal provinces are in the second and third groups, the left majorities are distributed in the fourth or fifth groups. By method of Multiple Linear Regression Analysis, the authors further analyses 6 impact factors with 14 variables (V₁-V₁₄). The result shows that human quality and information & knowledge productivity have stronger impact on Internet diffusion than the other 4 factors such as economic development level, opening up level, existing telecommunication infrastructure level and urbanization level. Stepwise Regression Analysis on the 14 variables produces that the standard partial regression coefficients of V₇, V₃,V₈,V₉,V₁₄ and V₁₂ are 0.709,0.595,-0.507, 0.490,-0.405 and 0.096 respectively,which indicates that number of patent per 10 thousand people, number of the high educated per 10 thousand people, share of the import and export value in GDP, hosted foreign tourists per 10 thousand people, urbanization level and regional density of laser cable have strong impact on Internet diffusion.