Abstract: Correlation between factors has an effect on the analysis of dam seepage observation data. Besides, the regression model is usually linear and it can hardly reflect the change of a dependent variable which is nonlinear function. In consideration of the above problems, and in combination with principal component analysis and BP neural network. A principal component analysis and BP neural network model is set up for analysis of dam seepage observation data in this paper. The result of an example shows that the prediction is more precise than that of the regression model.