A　new　method　for　noninvasive　detecting　coronary　artery　disease　(CAD)　based　on　BP　artificial　neural　network　(ANN)　is　presented,　and　the　diagnosis　system　is　designed.　The　ECG　signals　from　body　surface　are　proceeded,　and　four　character　parameters　(RR　interval,　width　of　QRS　complex,　scope　and　slope　of　ST　segment)　of　the　ECG,　correlating　with　CAD,　are　computed.　CAD　is　detected　and　diagnosed　by　using　BP-ANN,　which　is　imputed　with　character　parameters　of　the　ECG,　blood　pressure　and　basic　information　(sex,　age,　weight,　smoking　or　not).　Experiment　results　show　that　the　method　mentioned　in　this　paper　may　be　applied　and　easily　manipulated　for　CAD　diagnosis.