Background:The composite variability index (CVI), derived from the bispectral analysis (BIS), has been designed to detect nociception; however, there is no evidence that bilateral BIS and CVI show intrapatient reproducibility or variability.Methods:We conducted an observational study in patients who underwent for total knee arthroplasty. A BIS Bilateral Sensor was applied and continuously recorded at different points of the anesthesia procedure. Bland–Altman limits of agreement and dispersion for BIS and for CVI were applied.Results:Forty-nine right-handed patients were studied. There were differences between the right and left BIS values after tracheal intubation (which was higher on the right side) and at surgical stimulus (higher on the left side). The maximum BIS and minimum, mean, and maximum CVI scores were higher on the left side for left-side procedures, but there were no differences in any indexes for the right-side procedures. Except for the baseline measurements, both CVI and BIS scores presented high interpatient variability. Although the right to left bias was < 3% for the BIS index, dispersion was large at different stages of the anesthesia. The right to left bias for the CVI was 3.8% at tracheal intubation and 5.7% during surgical stimulus.Conclusions:Our results indicate that the large interindividual variability of BIS and CVI limits their usefulness. We found differences between the left and right measurements in a right-handed series of patients during surgical stimuli though they were not clinically relevant.