OBJECTIVE To establish a model of acoustically evoked short latency negative response (ASNR) in guinea pigs, a model of profound hearing loss with normal saccular functions, and verify the correlation between ASNR and vestibular evoked myogenic potential (VEMP). METHODS Thirty-two healthy guinea pigs were employed in the experiment, which were randomly divided into control group (16 subjects) and deafened group (16 subjects). Each animal experienced auditory and vestibular tests including auditory brainstem response (ABR), VEMP and caloric test. A quick treatment was employed for deafened group consisting of a subcutaneous injection of kanamycin at a dose of 400 mg/kg followed by a jugular vein injection of ethacrynic acid at a dose of 40 mg/kg one hour later. The animals were received ABR, VEMP and caloric test 7 - 10 days following the drug administration. The deafened group was further divided into ASNR group and non-ASNR group, based on the presence of ASNR. RESULTS In deafened group, five subjects died postoperatively, 11 subjects (22 ears) provided full data, ASNR was elicited in eight ears (36.4%), the threshold was 120 - 130 dB SPL with mean of (124.4 ± 4.96) dB SPL. Its latency range was 1.75 - 2.60 ms with mean of (2.15 ± 0.27) ms. The mean latency of threshold was (2.34 ± 0.18) ms. All eight ASNR ears presented with VEMP. The VEMP threshold, positive and negative potential latencies proved no statistical difference (P > 0.05) between ASNR group and control group. Significant difference was detected between the VEMP presence of ASNR group and non-ASNR group (P = 0.002). There was no statistically significant correlation between VEMP and caloric test neither between ASNR and caloric test in deafened group. CONCLUSIONS This study evoked ASNR in an ototoxicity guinea pig model which has profound hearing loss with normal saccular functions. The presence of ASNR correlated with VEMP, however, not correlated with caloric test, suggesting that ASNR and VEMP are both originated from the saccule.