The purpose of this study was to evaluate the performance of adults with traumatic brain injury (TBI) on a spoken persuasive discourse task and to evaluate the affects of eliciting this language sample. Ten adults with TBI (mean age = 51 years and 5 months) and ten adults matched by age and gender completed two spoken language tasks. These tasks required them to verbally provide their opinion of whether trained animals in circuses should be allowed to perform for the public and also whether public transport should be encouraged for everyone to use. One of the tasks was provided with examples for and against the topic within the instructions whereas the other task instructions provided no examples. The presentation of these tasks was alternated within the groups so as not to assist with task practice. Language measures included productivity (total number of words, mean length of T-units, T-units per minute and percentage of T-units with mazes) and complexity (total number of clauses, clause density and clause type). Pragmatic measures included the essential features of argument as identified in the developmental literature (number of claims, reasons, elaborations, repetitions, irrelevancies, and presence of an introduction and conclusion). The TBI group out-performed their age-matched peers on language complexity measures of total number of clauses and independent clauses used, however used significantly more adverbial clauses. On comparison of the elicitation technique, the instructions with examples elicited a significantly greater number of reasons than that of the basic instructions. The results are discussed alongside current literature in the field of discourse production and persuasion. Implications for clinical practice and future directions for research in this area are also suggested.