Abstract: In order to develop a high-quality Chinese precipitation product at the highly spatial temporal resolution, the optimum interpolation (OI) technique was adopted to combine the CMORPH with the Chinese Precipitation Analyses (CPA)，which was based on the hourly gauged rainfall from about 30,000 automatic weather stations (AWS), at 0.1°lat/lon resolution. The CMORPH was set to the first guess, while the CPA was used as observations to modify the first guess. The error statistics of the first guess and observations were defined based on the data from June to August of 2009 and the error structures were employed to compute the weights. The examination of the combined precipitation (COMB) at 00:00 UTC 2 July 2009 showed that the CMORPH at the target grid was able to be improved by the CPA if the gauge within a searching radius was available. Otherwise, the CMORPH was remained where there was no gauge within the searching radius. The independent validation result of the COMB from June to August of 2009 was that, the Bias/RMSE/RE of the COMB were -0.004 mm/h, 1.271 mm/h and 15.964%, respectively, with the averaged spacial correlation coefficient of 0.778. It was indicated that the COMB had a better quality than the CMORPH, as well as the FY-2C merged precipitation product.