The use of hand gestures during communication is a naturally occurring phenomenon thought to aid in cognitive processes such as language, memory, and learning. Co-speech gestures increase in individuals with language impairment after brain injury, e.g., post-stroke aphasia. Traumatic brain injury (TBI) provides a unique avenue to evaluate gestures, as individuals with TBI often present with both anomia and cognitive impairments, resulting in cognitive-communicative disorder. Broadly, this study examines the extent to which iconic gestures are present in procedural narratives post moderate-severe TBI, and their stability across a period of longitudinal recovery. This study has several aims: (1) identify longitudinal changes in overall iconic gesture frequency, (2) examine longitudinal iconicity representation, (3) examine the relationship between language assessment scores and iconic gesture use, and (4) quantify the use of iconic gestures that are supplemental to speech, i.e., add to, disambiguate, or replace speech. Archival audiovisual data was drawn from a previous longitudinal study investigating communication recovery in moderate-severe TBI. We found that gestures were employed consistently across 2 years. Additionally, we found a high proportion of iconicity types, which were also used consistently longitudinally. We found that participants did not employ many gestures that were supplemental to speech, however there is some evidence that under severe language impairment, gestures may be employed for conveying information that is not available in speech. This study extends previous TBI and aphasia gesture research, suggesting that gestures and speech are highly connected speech processes, and that gestures are consistent over time.