The invention discloses an assimilation method for satellite and ground rainfall measured values based on Kalman filtering. The assimilation method comprises the following steps of: pre-processing: acquiring satellite measured data and ground rainfall station data, which are mutually independent; meanwhile, carrying out comparative analysis on measured values to obtain measurement deviation; updating time: representing estimated values of the measured values and a noise error estimated value as self-correlation functions; and updating a method: assimilating an estimated value of the measurement deviation and an estimated value of noise error variance through a Kalman filtering method. With the adoption of the assimilation method, errors of the system are revised; noise influences can also be effectively eliminated; and the reliability of a rainfall estimated value is further improved.