Objective: To identify an optimal gaze angle for precisely evaluating the function of vertical semicircular canals in video head impulse test (vHIT) and draw a comparison on the performances among three different systems for vHIT. Design: Two-center prospective study. Setting: Participants were sit 1.2 meters away from the wall in a noise-proved room that dedicated for vHIT experiments. For the gaze-angle projects, targets were placed on the wall sequentially at the pre-marked lines for different angles. During the comparison experiments, similar settings were ensured in both hospitals, like distance to wall and angle of staring. Same examiner performed the comparison between two systems in one location. Participants: 26 healthy participants were recruited in the gaze-angle experiments, 16 of which were further involved in inter-examiner tests. For the comparison projects, 9 and 13 participants were recruited, respectively. Any participant with otologic or vestibular disorders was excluded. Results: Our research showed that instead of right ahead at 0 degree, when torsion applied at vertical semicircular canal planes, 25-degree is a better place to set the targets. At this angle, a more accurate VOR gain was easier to be obtained. Besides, our evaluations of three different systems showed that a new vHIT system, VertiGoggles ZT-VNG-I (VG) performed as good as the long-tested OtometricsICS impulse (Oto) and EyeSeeCam (ESC). Conclusion: We proposed a new protocol to set the targets 25 degrees from right ahead after tilt head 45 degrees to evaluate vertical canals during vHIT. Furthermore, the new VG system is good for clinical practices.