From 1989 through 1991, we conducted surveillance of spinal cord injury (SCI) among residents of Utah. We found an annual incidence rate of 4.3 per 100,000, with the highest rates occurring among males 15-24 years of age. Motor vehicles were the leading cause of injury, followed by falls, and sports and recreation. We also examined the accuracy and completeness of reporting in this surveillance system. We found the predictive value positive of SCI diagnoses reported in hospital discharge data to be only 61%. When we considered only patients who received acute hospital care in-state, we found that the sensitivity of hospital discharge data 89%. These findings indicate serious problems in the reporting of spinal cord injury diagnoses in hospital discharge data and the need to verify case reports based on these data. There is also a need to study this problem in other jurisdictions to determine if overreporting is widespread.