Abstract: In view of the deficiencies in the researching methods of urban distribution pattern in China, taking national garden counties as a special example, using methods of ESDA and imbalanced index to characterize the evolution of the number of provinces in the distribution pattern; Stepwise regression model, GWR and GeoDetector are also be applied to explore the influencing factors of its distribution and its spatial heterogeneity and interaction on the provincial scale. The results show that: (1)The distribution of national garden counties has been characterized by significant global and local spatial autocorrelation. The spatial heterogeneity of the counties has remained stable, but the number of the counties has become more discrete among provinces as a whole. (2)Total investment in fixed assets, green space ratio in built-up area of county, number of (Autonomous)Counties/Banners and local general public budget revenue are the significant factors leading to the spatial differentiation of the number of national garden counties. In case of the direction affected, these factors are all positive except for the last one. (3)The effects of each factor demonstrate obvious spatial non-stationarity, but the spatial distribution patterns of their effects are different; The influence is also not independent, there are two kinds of interaction, including nonlinear enhancement and bi-factor enhancement, the influence of which is greater than that of factors acting alone, and green space ratio in built-up area of county has the most promoting effect on other factors.