Local anaesthetics have been used intravenously and intratympanally to reduce tinnitus. In order to clarify its action in the periphery, we applied 0.5 mM tetracaine in the scala tympani in 18 cochleae and studied the effects on the receptor potentials. We used a temporal bone preparation of the guinea pig ear in vitro exposing the fourth cochlear turn where the cochlear microphonics (CM) and the summating potential (SP) were recorded. The perfusion was kept at a rate of 50 microliters/min. The frequency response of the cochlea was determined at the beginning of each experiment and the responses were recorded at the best frequency of the preparation. In another five cochleae an accumulated dose-response relationship was determined by increasing the tetracaine concentration in steps (50, 100, 300, 500, 1000 and 2000 microM), measuring the difference in amplitude of the receptor potentials. The CM decreased significantly (p < 0.001; mean 0.37 mV; SD 0.29). In 12 cochleae the SP was initially positive and did not increase significantly (p = 0.16; mean 0.07 mV; SD 0.16). In six cochleae the SP was initially negative and all changed polarity to positive and increased significantly (p < 0.05; mean 0.36 mV; SD 0.28). The effects on both the CM and the SP were reversible. Owing to the inter-individual variation between the cochleae the SP/CM ratio was determined and it increased significantly (p < 0.001; mean 0.18; SD 0.11). In the accumulated dose-response experiments the CM decreased significantly (p < 0.05) in a dose-dependent way, whereas the SP did not increase significantly. The SP/CM ratio increased significantly (p < 0.05) at 300 microM and 500 microM. We hypothesize that the peripheral tinnitus-reducing action of local anaesthetics is in part due to a reversal of the SP, but also to a reduction of the CM. The difference in effect of tetracaine on the receptor potentials, the CM and the SP, suggests that the SP is not dependent on the CM.