Aim: This exploratory study examined the relationships among different types of social support accessed by patients with traumatic spinal cord injury (SCI). Method: Using a cross-sectional study design, 20 SCI patients of workforce age discharged from the Royal Talbot Rehabilitation Hospital, Melbourne, Australia during 2007 were assessed on a range of demographic, injury and social support variables. Results: Higher quality interactions (i.e., actual social support) were significantly associated with greater social integration (i.e., structural social support) and higher perceived social support. However, community integration and social integration, which measure different aspects of structural social support, were essentially unrelated. Conclusions: Actual social support influences perceived social support and social integration. Theoretically, these results suggest that actual social support influences the two main opposing models that explain the relationship between social support and wellbeing: the stress-buffering model (for perceived social support) and the main-effects model (for structural social support). From a research perspective, the use of well-defined types of social support, and psychometrically sound measures, is critical for assessing the influence of social support on rehabilitation outcomes. To effectively match rehabilitation efforts with SCI survivors' needs, future research should examine which types of social support best predict particular rehabilitation outcomes.