This study assessed the test-retest reliability/repeatability and 95 percent confidence intervals (CIs) of the Tinnitus Handicap Inventory (THI) and developed categories for classifying self-perceived tinnitus handicap severity. Twenty-nine adults with tinnitus as their primary auditory complaint served as subjects. The THI was administered on two occasions (mean interval 20 days) using a paper-pencil format. Results support the contention that the THI is psychometrically robust. Specifically, the test-retest reliability/repeatability was high. Additionally, the 95 percent CI for the THI was 20-points, indicating that a 20-point or greater change had to occur from test to retest for a change to be considered statistically significant at the 5 percent confidence level. Quartiles calculated from raw scores were used to create a matrix of values representing tinnitus severity. We conclude that the THI is a brief, easily administered, and psychometrically robust measure that evaluates the impact of tinnitus on daily living.