Abstract: Aiming at the mutation characteristics of R-wave of ECG signal, a method of R wave detection based on adaptive cell averaging-constant false alarm rate (CA-CFAR) is proposed by using radar signal detection method. Firstly, filter banks are used to preprocess ECG signals; then, adaptive CA-CFAR is used to detect and judge the preprocessed signals; finally, a rule of refractory elimination is processed according to the interval characteristics of R waves of ECG signals, and the location of R waves is obtained. The simulation of electrocardiogram (ECG) signals in MIT-BIH database provided by MIT shows that the accuracy of R-wave detection of ECG signals in MIT-BIH by CA-CFAR of adaptive reference unit is 99.842% and the detection error is 0.354%. The measured data show the effectiveness and applicability of the algorithm.