To assimilate wind profiler data into operational numerical weather model, this paper analyzes the wind profiler product quality characteristic for data assimilation. Quality analysis is the basic work of this kind of data quality control and assimilation. Therefore, this paper analyzes the quality characteristics of the wind profile radar wind product through comparing the data and numerical model background field, and using the sounding and rainfall data. And the wind product used in this article is from the Meteorological Data Storage System (MDSS) in National Meteorological Information Centre. The findings show that average hourly sample wind is better than realtime sample wind for data assimilation application. There are lots of wind profile radar wind products existing above the effective detection height. And the bias of these data against numerical background is abnormally large. The bias between wind profiler and model background for PA Mode radar is the smallest, the second smallest is that for PB Mode and the biggest bias is for LC Mode within the effective detection height range, where there is no bias between wind profiler data and the numerical prediction background. The amount of data under one standard deviation is more than 89%, two standard deviations of the data reach more than 97%, three standard deviations of the data is above 98%. Meanwhile, precipitation is an important factor affecting the data quality, so when applying the radar product, the precipictation data should be analyzed carefully for as quality contiol.