Abstract: Using social network analysis method, the paper studies the geographical distribution and network structure of self-driving tourist flow in Fujian province. It can be concluded as follows from the market characteristic. (1) Peak and trough of tourist season alternate four times in a year, showing M model, especially peak seasons achieve a high level compliance with Chinses long holiday. (2) The travel models mainly between couples , parent-child and parents account for 60.1%，secondly relatives and friends accounting for 27.1%. (3)88.18% self-driving tourists stay in Fujian province less than 6 days, of which 3 days accounting for 23.15%. (4) Self-driving tourist market present an adjacent character, a clear character for eastbound development and the character concentrated on the cities.(5) The tourism special use curve of self-driving is consistent with curves of Boltzman. The number of tourists increase with distance and peaks at 800km, then the number of tourist shows a rapid decrease. On the whole, its gravity field concentrates on the scope of 1400km, and the accumulation percentage of tourists reaches 87%. Spatial characteristic of self-driving tourist network into Fujian province are shown in the following: (1) The status of network node present the inequality, the polarization between core nodes and periphery ones has been clear. (2) The spatial distribution of nodes is featured by weak concentration and strong dispersion, while the nodes of core region and primary region account for less. (3)On the whole, it is quite clear of the self-driving tourist flow hierarchical order , the self-driving tourist flow concentrates on higher scenic which located such as Siming District, Nanjing County, and Gulou District etc.