Hierarchical Bayesian method is increasingly being used by ecologists. Methods to accomplish such predictions could explain uncertainties in observation, sampling, models, and parameters. Soil nitrogen model was built using Hierarchical Bayesian method that accommodates uncertainties of data and model provides a richer understanding of the model in Badaling region. At the same time, soil nitrogen content was predicted in different soil layers(A, B, C). The results show that: (1) Soil nitrogen modeling is yi~N(β0j\[i\],k\[j\]+β1j\[i\],k\[j\]xi,σ2y) for the research area. (2) Uncertainties of data and model indicated that the model is good to predict soil nitrogen content. (3) Prediction of soil nitrogen content showed that soil nitrogen content of A layer was increased with increasing of elevation. It was found that soil nitrogen content of plant type 0,1,2,3 were increased with the increase of elevation in B layer, however, soil nitrogen content of plant type 4 was decreased with the increase of elevation in the layer. With the increase of elevation, there was a increase observed in soil nitrogen of the vegetation type 0, while a decrease in that of vegetation types 1,2,3,4 in layer C. Soil nitrogen content of layer A was the greatest, followed by layer B and layer C. The result indicated that soil nutrient content decreased with increasing depth.