It is becoming increasingly important to make timely and accurate prediction of human blood pressure changes in order to prevent the exacerbation caused by the instable human blood pressure .This paper proposes a prediction model of human blood pressure based on the combination of wavelet analysis and BP neural networks. This model uses the wavelet decomposition and reconstruction method to decompose and reconstruct non-stationary human blood pressure sequence, separating the high frequency components and the low frequency components in the original sequence, then the BP neural network prediction algorithm is used to establish the prediction model for each layer. Finally, the predicted values of the two models are added to obtain the predicted values of the original series. The results show that the prediction accuracy of the combined forecasting model is obviously higher than that of the traditional BP neural network prediction model, which provides an effective and reliable combination forecasting method for human blood pressure prediction.