This study applied the vestibular evoked myogenic potential (VEMP) test to guinea pigs coupled with electronic microscopic examination to determine whether VEMPs are dependent on type I or II hair cell activity of the saccular macula. An amount of 0.05 ml of gentamicin (40 mg/ml) was injected directly overlaying, but not through, the round window membrane of the left ear in guinea pigs.One week after surgery, auditory brainstem response test revealed normal responses in 12 animals (80%), and elevated thresholds in 3 animals (20%). The VEMP test using click stimulation showed absent responses in all 15 animals (100%). Another 6 gentamicin-treated animals underwent the VEMP test using galvanic stimulation and all 6 also displayed absent responses. Ultrathin sections of the saccular macula in the gentamicin-treated ears displayed morphologic alterations in type I or II hair cells, including shrinkage and/or vacuolization in the cytoplasm, increased electron density of the cytoplasm and nuclear chromatin, and cellular lucency. However, extrusion degeneration was rare and only present in type II hair cells. Quantitative analysis demonstrated that the histological density of intact type I hair cells was 1.1 +/- 1.2/4000 microm(2) in the gentamicin-treated ears, showing significantly less than that in control ears (4.5 +/- 1.8/4000 microm(2)). However, no significant difference was observed in the densities of intact type II hair cells and supporting cells between treated and control ears. Furthermore, the calyx terminals surrounding the damaged type I hair cells were swollen and disrupted, while the button afferents contacting the damaged type II hair cells were not obviously deformed. Based on the above results, we therefore conclude that VEMPs are heavily dependent on type I hair cell activity of the saccular macula in guinea pigs.