Major sources of data on incidence and prevalence of spinal cord injury (SCI) in the United States were reviewed, including many federal government agencies, state level rehabilitation and health departments, regional SCI systems, private voluntary organizations, available literature, and several unpublished studies. With the exception of the Veterans Administration, many of the federal databases did not prove helpful. For example, the estimates of the annual surveys conducted by the National Center for Health Statistics (NCHS) of traumatic SCI are likely to be inflated due to multiple hospital admissions of the same patient and deflated due to misclassification of SCI. The tendency for inflation and deflation, however, may balance each other, resulting in a rough estimate of annual traumatic SCI incidence of about 11,000 for 1984. Data from the Florida and Virginia Central Registries of SCI suggest a decrease in traumatic SCI incidence in recent years. Based on data abstracted from a regional study, the California Disability Survey, we estimate the prevalence of traumatic SCI to be about 238,000 in the United States for 1984.