A model was developed to provide a tool to forecast demographic trends in populations of people with traumatic spinal cord injury at the national and state level. This information is critical to planning for the allocation and distribution of resources to care for people with spinal cord injury. The literature on incidence, mortality, and prevalence of spinal cord injury in the United States was reviewed and reported values were evaluated for incorporation into the model. A linear relationship between age specific survival rates of the spinal cord injury population, and expected survival rates in the absence of spinal cord injury was established and this provided the basis for projections using age cohort survival methodology. The model's projections indicate a need for future expansion of capacity to treat traumatic spinal cord injury in the private sector, and a need to prepare for an aging disabled population. The annual number of traumatic spinal cord injury cases admitted to hospitals is projected to increase from approximately 11,500 in 1994 to almost 13,400 in 2010. Age adjusted post-hospitalization incidence rate in 1994 is estimated at approximately 38 per million (23 per million for females and 55 per million for males). A 20% increase in the US spinal cord injury prevalence can be expected over the next 10 years, going from approximately 207,000 estimated in 1994, to 247,000. During this time, the veteran segment, which currently comprises 22% of the spinal cord injury population, is projected to decline.(ABSTRACT TRUNCATED AT 250 WORDS)