To develop a mouse model for temporary and persistent tinnitus using the gap startle paradigm.Behavioral animal models for tinnitus are classified into conditioning- and reflex-based types. Gap prepulse inhibition of the acoustic startle (GPIAS) is based on the acoustic startle modification by a silent gap and gap detection deficit caused by tinnitus gap filling.We used C57BL/6J mice inherently susceptible to hearing loss and potentially predisposed to tinnitus. They were divided into the control, salicylate-induced tinnitus, and noise-induced tinnitus groups. Mice were tested with the auditory brainstem response at four frequencies (8, 16, 24, and 32 kHz) and GPIAS in three carrier conditions, 16 and 24 kHz narrow band noises (NBNs) and broadband noise, at multiple time points before and after treatment. The ratio between the gap startle and no-gap startle amplitudes was analyzed by a repeated measures design. In addition, the number of tinnitus-positive mice meeting a specified criterion was counted.Salicylate/unilateral noise trauma resulted in temporary/permanent tinnitus evidenced by GPIAS reduction. GPIAS reduction was the most significant at 16 kHz NBN among the three carriers in both tinnitus groups. Control mice also showed good gap detection performance at 16 kHz NBN, which is in the most sensitive hearing range in mice.The GPIAS test in C57BL/6J mice was very reliable at 16 kHz NBN. This tinnitus model developed in the mouse strain of accelerated hearing loss can be used with two options of temporary and persistent tinnitus.