In the decisive stage of developing of a moderately prosperous society in all aspects, hazy weather has become a major obstacle to the further advancement of China. Therefore, improving the level and efficiency of haze governance has become essential. Based on the DPSIR model, this paper builds a haze governance level and efficiency index system using the entropy method and the super-efficiency data envelope-analysis (DEA) model to analyze the data for 31 cities in North China from 2007 to 2016. From the aspects of spatial differences and influence factors influencing the comparative analysis, the results are as follows. (1) During the investigation period, the level and efficiency of city haze governance in North China showed a trend of fluctuation and decline, with obvious stages in their characteristics. Haze governance efficiency is much higher than its level, and its mean value reaches the DEA level which indicates that it is effective. (2) A significant regional gradient difference occurs between these two aspects. The haze governance level presents a convex distribution pattern of “east low–middle high–west low”, while the haze governance efficiency presents a concave distribution pattern of “east high–middle low–west high”. (3) The regression results show that economic growth has a negative effect on both haze governance level and efficiency. By contrast, the industrial structure has a positive effect on haze governance level and efficiency, but the significance of its effect on these two is different. On this basis, policy suggestions are proposed for improving the level and efficiency of haze governance in various cities in North China.