Objective: Tinnitus assessment and outcome measurement are complex, as tinnitus is a purely subjective phenomenon. Instruments used for the outcome measurement of tinnitus in the context of clinical trials include self-report questionnaires, visual analogue or numeric rating scales and psychoacoustic measurements of tinnitus loudness. For the evaluation of therapeutic interventions, it is critical to know which changes in outcome measurement instruments can be considered as clinically relevant. For this purpose, the concept of the minimal clinically important difference (MCID) has been introduced. Study design: Here we performed a literature research in PubMed in order to identify for which tinnitus outcome measurements MCID criteria have been estimated and which of these estimates fulfil the current methodological standards and can thus be considered as established. Results: For most, but not all tinnitus outcome instruments, MCID calculations have been performed. The MCIDs for the Tinnitus Handicap Inventory (THI), the Tinnitus Questionnaire (TQ), the Tinnitus Functional Index (TFI) and visual analogue scales (VAS) vary considerably across studies. Psychoacoustic assessments of tinnitus such as loudness matching have not shown sufficient reliability and validity for the use as an outcome measurement. Conclusion: Future research should aim at the confirmation of the available estimates in large samples involving various therapeutic interventions and under the consideration of time intervals and baseline values. As a rule of thumb, an improvement of about 15% can be considered clinically meaningful, analogous to what has been seen in other entirely subjective pathologies like chronic pain.