By using spatial hot spot analysis and spatial error regression model, the authors analyzed the spatial distribution pattern and the influencing factors of unsafe charging behavior for e-bikes in Guangming district of Shenzhen, with the daily inspection data collected in urban grid management. Urban spatial features, such as land use types, employment opportunities, facility and traffic conditions were found to have a close correlation with unsafe charging behaviors at different significance levels. There was a certain spatial autocorrelation of the unsafe charging behavior of e-bikes. Most of the unsafe charging concentrated in a few grids, and 9.2% of the grids in the case area covered 84.94% of the total number of unsafe charging behaviors. Variables related to travel demand and traffic conditions have a stimulating effect on people's behavior of illegal charging of e-bikes. There was a positive correlation between the proportion of residential land, the number of enterprises and the road density within the urban grid and the unsafe charging behaviors of e-bikes. The standardized regression coefficients were 0.09, 0.03 and 0.02 respectively. There was a negative correlation between the proportion of public facilities and public building land in urban grid, the distance from home to primary schools and the unsafe charging behaviors of ebikes, and the standardized regression coefficients were ?0.02 and ?0.01. Higher coverage rate of bus stop would reduce the use of e-bikes, and lead to less unsafe charging behaviors. The results can be used to direct the optimization of patrol routes in grid management and improve the ability of pre-perception and fine prevention of urban fire accidents.