The purpose of the present study was to investigate the ability of three different normalization methods, which are mean value, standard deviation and root mean square (RMS) value, based on photoplethysmographic pulse beat interval (PBI) in response to skin incision under various remifentanil concentrations during general anaesthesia. Forty-one female patients who had elective gynaecological surgery with propofol-remifentanil target-controlled anaesthesia were randomly allocated to one of four different remifentanil effect-site concentrations (0, 1, 3 and 5 ng•ml −1 ), Finger photoplethysmography (PPG) waveforms were recorded for two minutes before and after skin incision, respectively, and PBI was extracted off-line. Three normalized measures PBI mean , PBI std , and PBI RMS were compared with the PBI values without normalization. Two patients were excluded from the study due to data missing or poor data quality. Before skin incision, only PBI without normalization was adversely affected by remifentanil concentrations. After the skin incision, PBI mean , PBI std , and PBI RMS performed better than PBI in discriminating the insufficient concentrations (0 and 1 ng•ml −1 ) from sufficient concentrations (3 and 5 ng•ml −1 ) of remifentanil. PBI mean was superior to all other variables in distinguishing the sufficiency of analgesia, with the largest area under the curve of the receiver-operating characteristic curve $(\text{AUC}=0.874$). These results demonstrated that normalized PBIs were less affected by changes in remifentanil concentrations that were unrelated to the analgesic-nociceptive balance. Compared with standard deviation and RMS value, Normalization with mean value is a better approach for pre-processing of PBI. Further studies of mean value for normalization are needed to elucidate its usefulness during other anaesthetic and surgical conditions.