This paper presents on breast cancer knowledge mining and network to its effective to explore. Before the application of various data mining method in probability, find out the progression of the disease by using the network. The results are helpful to reasonable treatment to the doctor for the patient. For the high dimension data set and overcome the problems associated with the data, this paper uses principal component to reduce the data dimension and find out the suitable network analysis. Application of BP neural network structure is assessed, performance on the breast cancer data in a hospital in Beijing, the BP neural network is analyzed. The results show that principal component analysis to eliminate the correlation between the input of the network, reducing the input of network, improve the performance of the network as a whole. The good results have been achieved.