OBJECTIVE: Tinnitus is a distressing symptom for which few treatments exist. It leads to an important decrease in quality of life in 2 to 3% of the population. Tinnitus is considered a phantom sound, the result of cortical reorganization. Transcranial magnetic stimulation (TMS) is a noninvasive method to modulate cortical reorganization and has been shown to be able to influence tinnitus perception. STUDY DESIGN: Retrospective analysis. SETTING: Tertiary referral center. PATIENTS: The effect of TMS of the contralateral auditory cortex in 114 patients with unilateral tinnitus is investigated as one of the selection criteria used for surgical implantation of electrodes on the auditory cortex. INTERVENTION: TMS is performed at 90% of motor threshold at 1, 3, 5, 10, and 20 Hz, with each stimulation session consisting of 200 pulses. Results were classified as no effect (0-19% improvement), partial effect (20-79% improvement), and good effect (80-100 suppression). MAIN OUTCOME MEASURES: TMS had a good effect in 25% of the patients studied, partial effect in 28% patients, and no effect in 47%. RESULTS: TMS at 200 pulses is capable of tinnitus suppression for seconds only. The results were influenced by tinnitus duration: the longer the tinnitus exists, the lower the stimulation frequency that yields maximal tinnitus suppression (p < 0.001). The maximal amount of tinnitus suppression decreases in time (p < 0.01), resulting in a 2% decrease of potential tinnitus suppression per year. CONCLUSION: TMS of the auditory cortex is capable of modifying tinnitus perception for a very short time. The maximal amount of suppression and best stimulation frequency depends on the tinnitus duration.