Abstract: In order to solve the problem of data loss in mapping paddy fields, a novel rice extraction method was proposed in this paper by using spatial and temporal remote sensing data fusion technology (STDFT). Using the temporal change information extracted from time series MODIS data, combining with the texture information of Landsat-ETM+ data at former, fusion data was produced which had the same temporal resolution with MODIS data and the same spatial resolution with Landsat-ETM+ image. Then using fusion data in critical period, rice fields was mapped through SVM. The algorithm had been tested over a study area in Jiangning Country, Nanjing City, Jiangsu Province, China. Results showed that this method could map rice effectively. High mapping precision of 93% was acquired with Kappa coefficient of 0.96