To explore the energy proportion of the QRS complexes in the electrocardiogram (ECG), a new method, wavelet analysis, has been presented. Using wavelet transform of ECG, we can calculate the energy proportion of QRS complexes in multiple scales. The electrocardiograms (ECGs) were used as the experimental data, which were collected from a young (21-34 yr) group and an elderly (68-81 yr) group of healthy subjects, as well as from a group of arrhythmia patients (66-81 yr). the data analysis was performed with the energy proportion of the QRS complexes in the ECG using Mexican-Hat as a mother wavelet in multiple scales. Results showed that the energy proportion of the QRS complexes had no changes with ages increasing (P > 0.44), but in the same age group, the arrhythmia patients' energy proportion of the QRS complexes near 17Hz are obviously less than that in the healthy group (P < 0.01), so the energy proportion of the QRS complexes calculated by wavelet analysis can be used as a feature index to judge whether a person is a sinus arrhythmia patient or not.