OBJECTIVE: There is a need to develop and validate a brief questionnaire to measure tinnitus severity that can meet the needs of clinicians and researchers. The objective of this study is to demonstrate the validity and reliability of the 9-item Tinnitus Severity Short Form (TS-SF). STUDY DESIGN: A cohort study of online survey respondents recruited in October 2020 from the American Tinnitus Association electronic mailing list with a 2-week follow-up time interval. SETTING: Online. METHODS: Face and content validity, factor analysis, internal consistency, convergent validity, reliability, and minimal clinically important difference (MCID) were analyzed. RESULTS: Of 1121 participants that started the survey, 325 (29%) completed the baseline TS-SF. Through confirmatory factor analysis, the 9-item survey demonstrated a bifactor model with a substantive distress factor and a method factor. The instrument had high internal consistency with Cronbach's α of .96. Convergent validity determined the TS-SF score to be a predictor of tinnitus severity. Of 261 follow-up respondents, the test-retest reliability of r = 0.69 (95% confidence interval, 0.62-0.75) was determined. The MCID was determined using a distribution approach to be a 12-point change. CONCLUSION: The TS-SF demonstrates good internal consistency, validity, and reliability. This patient-reported outcome measure for tinnitus can be used in tinnitus treatment trials and clinical practice. The 9-item survey is less likely to result in respondent survey fatigue than longer surveys, and the 2-week reference interval allows for monitoring of tinnitus severity change in an appropriate timeframe.