In this study, we evaluated the psychometric properties of the Depression Anxiety Stress Scales 21 items (DASS-21) in a mild traumatic brain injury (mTBI) sample. Treatment-seeking adults (n = 347) were recruited from outpatient rehabilitation services in New Zealand. Dimensionality, reliability, person separation index, and differential item functioning (DIF) of the DASS-21 were examined using Rasch analysis. Initial analysis of the complete 21-item DASS showed poor overall fit due to problems with individual items. Fit to the Rasch model was excellent when treated as three composite scores. The stress subscale demonstrated adequate model fit, dimensionality and good reliability. For anxiety, fit was not good, reliability was unsatisfactory and DIF was evident on one item. When this item was removed, fit to the model was still inadequate as was reliability. DIF was also evident for depression, but when this item was removed, fit to the model was adequate. The DASS-21 is a psychometrically sound measure of distress and stress for adults seeking treatment following mTBI. Ordinal to interval score conversion tables are provided to increase the precision of measurement. When assessing depression in a mTBI population, a 6-item depression subscale is recommended. Caution is advised in using the DASS-21 anxiety subscale alone.