The aim of this study was to investigate the effect of alcohol on sacculocollic and vestibulo-ocular reflex systems, when the breath alcohol concentration (BrAC) is close to the legal limit of 0.25 mg/l. Twenty healthy male volunteers underwent vestibular evoked myogenic potential and caloric coupled with visual suppression tests. These tests were conducted prior to imbibing alcohol at a dosage of 0.5 g/kg to achieve a peak BrAC of around 0.25 mg/l. Once the peak BrAC was reached, these tests were performed again. Predosing and postdosing analytical results were compared, as were those with BrAC levels ≧ 0.25 mg/l and <0.25 mg/l. After ingesting alcohol, 36 ears (90%) showed vestibular evoked myogenic potential responses, with a significantly increased latency of peak p13. The mean slow-phase velocity of caloric nystagmus in 40 ears after dosing was significantly reduced, and that with BrAC ≧0.25 mg/l was significantly less than that with BrAC <0.25 mg/l. Likewise, the visual suppression index decreased considerably after alcohol ingestion. In conclusion, from the perspective of vestibular function, the 0.25-mg/l BrAC limit gains clinical significance, because the vestibulo-ocular reflex performance deteriorates further, when the BrAC exceeds 0.25 mg/l. However, impaired performance of sacculocollic reflex and vestibulocerebellar interaction has occurred, when the BrAC was <0.25 mg/l, suggesting that a lower legal threshold is appropriate.