Abstract Objective: To appraise the added benefit of refixation saccades (RS) towards the improvement of the video head-impulse test (vHIT) diagnostic accuracy in cases of suspected left horizontal semicircular canal dysfunction. Study Design: Case-control. Setting: Tertiary referral center. Participants: Twenty patients with a final diagnosis of left horizontal semicircular canal dysfunction and 20 patients for whom vestibular dysfunction was ruled out. Intervention: vHIT recordings of 40 patients with left horizontal semicircular canal (LHSCC) vestibulo-ocular reflex (VOR) gain <0.8. Main outcome measures: LHSCC VOR gain; Presence of RS and their frequency, latency, and velocity characteristics. Results: Gain values > 0.72 were found in all patients with no vestibular disease and in 4 (20%) patients having vestibulopathy. Significantly higher average left-sided RS velocity and frequency were found among the vestibular patients. VOR gain < 0.72 was found to be highly specific for the diagnosis of vestibular dysfunction. However, for gain values in the range of 0.72-0.79 the presence of RS with frequency > 80% largely improved vHIT diagnostic accuracy. CONCLUSIONS: Although VOR gain<0.8 is considered to reflect dysfunction a significant false positive rate for left-sided horizontal vHIT was found for gains in the range of 0.72-0.79. The presence of RS with frequency >80% could improve vHIT diagnostic ability in these patients. Key words: video head impulse test, re-fixation saccades, peripheral vestibular pathology, gain asymmetry in vHIT, lateral semicircular canal hypofunction, saccadic frequency, corrective saccade velocity