Spatial panel data models have been widely studied and applied in both scientific and social science disciplines, especially in the analysis of spatial influence. In this paper, we consider the spatial dynamic nonparametric Durbin model (SDNDM) with fixed effects, which takes the nonlinear factors into account base on the spatial dynamic panel data models (SDPDM). Above all, we propose an iterative approach to estimate the spatial dynamic nonparametric Durbin model with corresponding hypothesis test and we find that convergence occurs since the second iteration. So we use a three stage iterative approach to improve the iterative approach. The results indicate that the three stage iterative approach is more reliable when T or N is large and the accuracy of the nonparametric components estimation is very important. We believe that the three stage iterative approach can be applied to other spatial dynamic nonparametric panel data models as well.