Based on the daily precipitation from observation stations in North China from 1961 to 2010, six extreme precipitation indexes are calculated and their spatial and temporal variation characteristics are analyzed. The simulated capability of global climate models which from CMIP5 on the extreme precipitation are further evaluated. The results are as follows: （1）The extreme precipitation indexes in North China show large regional differences. The values are high in south and low in north of North China for precipitation intensity and extreme precipitation thresholds. The rain days and continuous dry days are different between east and west of North China. The interannual variability of extreme precipitation indexes has a drying trend, and the intensity of extreme events is increasing at the same time. （2）There are differences between the global climate models of CMIP5 for simulating North China extreme precipitation index. The spatial distribution of precipitation intensity and extreme precipitation threshold have a better performance in the models. The pattern correlation coefficients are above 06 for more than half models. The same model has different simulated capabilities for different extreme precipitation indexes. （3） The pattern correlation coefficient and root mean square error have good consistency, and the correlation coefficient of model ranking is larger than 04 HadGEM2CC，HadGEM2ES，MIROC4h，MIROC5, and CCSM4 are the best five global models. Compared to allmodal ensemble average, the bestmodel ensemble average can significantly reduce the wet deviation of simulation.