Deficits in social functioning are common following traumatic brain injury (TBI). Previous research suggests that there are multiple, complex factors which underlie such deficits, including cognition and social communication. Standardized measurement of social communication following TBI is challenging, and often involves the use of structured rating scales. The Profile of Pragmatic Impairment in Communication (PPIC) is one such scale that shows promise, though it requires further development and empirical testing to improve its utility in the TBI population. The current study is archival in nature. Data were obtained from two prior study samples in projects investigating social communication in the community dwelling adult TBI population: the social communication assessment measures study (SCA study, N = 121) and a randomized clinical trial for social communication intervention (IPR study, N = 83). In order to further develop the PPIC and examine the underlying cognitive abilities which impact social pragmatics, the 84 behavioral items of the PPIC were reduced to a set of 20 items deemed to be most characteristic of social communication difficulties following TBI. These 20 items were analyzed using exploratory factor analysis. Following an iterative process, a two factor solution accounting for 60.77% of the total variance was obtained, and it included 9 of the 20 originally selected items. These factors were labeled Partner Sensitivity (5 items) and Conversational Flow (4 items), and subscale scores were created by summing the item scores within each factor. The cognitive underpinnings of social pragmatics as measured by the new PPIC subscales were examined using hierarchical linear regression, using measures of attention, executive functioning, and affect perception as predictor variables and the new PPIC subscale scores, AIPSS Overall Sending score, and the TIRR Social Communication Rating Form (an experimental measure) as outcome variables. After adjusting for demographic and injury-related variables, performance on cognitive measures accounted for a unique 22% of the variance in PPIC Conversational Flow scores and 17% of the variance in AIPSS Overall Sending scores, while performance on cognitive measures did not account for a statistically significant amount of unique variance in PPIC Partner Sensitivity scores or TIRR Social Communication Rating Form scores. These results represent important preliminary steps in the development of the PPIC into a more parsimonious and useful tool and in developing a more sophisticated understanding of the relationship between cognition, social communication, and social functioning in TBI. Advisors/Committee Members: Struchen, Margaret A (advisor), Hannay, Hilda J (advisor), Massman, Paul J (committee member), Steinberg, Lynne (committee member), Pappadis, Monique (committee member).