Abstract: Ensemble Kalman Filer (EnKF) is a flexible and effective sequential data assimilation method. Appling EnKF to solve the parameter optimization problem has several advantages. Firstly, it explicitly considers multiple sources of uncertainty, thus avoids the excessive adjustment of parameters due to the compensation for errors arising from other sources, which will generate sub-optimal parameters. Secondly, it processes the latest updated observations, eliminating the storage and processing all the historical data simultaneously. Thirdly, it characterizes and predicts related error statistics by Monte Carlo method, thus no closed solution approach is needed, making it easy to build with existing numerical models. In this study, by simulation of observed data, EnKF is evaluated in terms of the effectiveness and efficiency of calibrating soil hydraulic function parameters in one-dimensional Richards equa-tion. The results show that, optimal parameter estimates can be easier obtained for more sensitive parameters. The performance is influenced by neither the initial surmises nor the observation error settings. Contrary to intuition, the increase of assimilation frequency may cause instable estimation results.