In this paper, we explore the application of extended Fourier amplitude sensitivity test (EFAST) to land surface parameter sensitivity analysis. Based on observations of Tongyu site (Jilin province) at a degraded grassland in 2008, the EFAST method is verified and applied. The sensible heat flux and latent heat flux are used as the test variables. With the full consideration of the sensitivity of individual parameters, the coupling between the parameters is taken into account. We explore the influence on the nonlinear system under the constraints of the interaction of multiple land surface parameters with a quantitative analysis. The results show that sand content in the soil (sand) and minimal porosity of permeable water are the key factors which affect the heat sensible flux and latent heat flux significantly, which is consistent with the existing result. The sensitivity result confirms the feasibility of the EFAST method. The results of this paper are expected to guide people in designing the field observation and development of parameterization schemes in land surface model.