The measurement of anesthetic depth is important in anesthesiology. Although heart rate variability (HRV) is profoundly affected by general anesthesia, it has not yet been commonly used in this field. One of the reasons is the lack of suitable parameters of HRV for short-term observations. In this study, we designed a time domain parameter of HRV named the similarity index. It was based on observing the trend of the distribution of instantaneous heart rates as time moved. Taking epochs of ECG data as short as 64 s can derive the index. We observed the values of this index of 30 patients when they were awake and under isoflurane anesthesia. The values had very little overlapping between the two states and the prediction probability to distinguish the two states was 0.91. We suggest that HRV, if suitably treated, can play more roles in the monitoring of anesthetic depth.