In order to further promote the overall planning and coordination configuration of arable land in Beijing-Tianjin-Hebei region, the space control and comprehensive upgrading mechanism is implemented about arable land quality, and the demand of coordinated development of arable land resource needs to be ensured. Fractal characteristics of spatial distribution of arable land quality in Beijing-Tianjin-Hebei region and its mechanism were studied by the combining means of GIS, fractal theory and grey relational grade analysis. Main conclusions can be drawn as: in Beijing-Tianjin-Hebei region, the quality of arable land showed a trend with certain regularity in space with the characteristics of obvious differences between the north and the south, which was high in south and low in north, low in regional edge, and high in middle region. There was less arable land but good quality in Beijing and Tianjin and a large number of arable land in Hebei; but the quality was generally not high. The fractal theory can be used to study the structural characteristics of the spatial distribution of arable land quality, fractal features represented the spatial form complexity and spatial occupation of arable land quality in Beijing-Tianjin-Hebei region. The spatial structure of Beijing-Tianjin-Hebei region was the simplest with the most stable form and the highest quality, and its space occupying was the highest at the natural quality level. While at the economic quality level it was remained the lowest; the spatial structure and stability of medium land were in the middle, but the space occupying was the largest; spatial structure complexity, the stability of space form and the space occupying of low quality arable land were opposite to the higher quality arable land. The main factors that affected the fractal characteristics of arable land quality were the natural conditions, such as light, temperature and climate, based on this, agricultural production conditions and socio-economic conditions disturbed its fractal characteristics further.