Background: Personalization of repetitive transcranial magnetic stimulation (rTMS) for tinnitus might be capable to overcome the heterogeneity of treatment responses. The assessment of loudness changes after short rTMS protocols in test sessions has been proposed as a strategy to identify the best protocol for the daily treatment application. However, the therapeutic advantages of this approach are currently not clear. The present study was designed to further investigate the feasibility and clinical efficacy of personalized rTMS as compared to a standardized rTMS protocol used for tinnitus. Methods: RTMS personalization was conducted via test sessions and reliable, sham-superior responses respectively short-term reductions in tinnitus loudness following active rTMS protocols (1, 10, 20 Hz, each 200 pulses) applied over the left and right temporal cortex. Twenty pulses at a frequency of 0.1 Hz served as a control condition (sham). In case of a response, patients were randomly allocated to ten treatment sessions of either personalized rTMS (2000 pulses with the site and frequency producing the most pronounced loudness reduction during test sessions) or standard rTMS (1 Hz, 2000 pulses left temporal cortex). Those participants who did not show a response during the test sessions received the standard protocol as well. Results: The study was terminated prematurely after 22 patients (instead of 50 planned) as the number of test session responders was much lower than expected (27% instead of 50%). Statistical evaluation of changes in metric tinnitus variables and treatment responses indicated only numerical, but not statistical superiority for personalized rTMS compared to standard treatment. Conclusions: The current stage of investigation does not allow for a clear conclusion about the therapeutic advantages of personalized rTMS for tinnitus based on test session responses. The feasibility of this approach is primarily limited by the low test session response rate.