PURPOSE: The primary objective was to study factors influencing post-injury employment and withdrawal from work in persons who sustained traumatic spinal cord injury (SCI) more than 20 years ago. A secondary objective was to study life satisfaction in the same patients. METHOD: A cross-sectional study with retrospective data of 165 SCI-patients admitted to Sunnaas Rehabilitation Hospital 1961-1982. Multiple logistic regression was used to identify predictors for obtaining work post-injury. A Cox proportional hazards regression model was used to study factors influencing early withdrawal from work, i.e. time from injury until discontinuing employment. RESULTS: Sixty-five percent of the participants were employed at some point after the injury. Thirty-five percent still had work at the time of the survey. The odds of obtaining work after injury were higher in persons of younger age at injury, higher in males versus females, higher for persons with paraplegia versus tetraplegia, and for persons classified as Frankel D-E compared to a more severe SCI. Factors associated with shorter time from injury until discontinuing employment were higher age at injury, incidence of injury after 1975 versus before, and a history of pre-injury medical condition(s). Life satisfaction was better for currently employed participants. CONCLUSION: The study indicates a low employment-rate in persons with SCI, even several years after injury. From the results, we suggest more support, especially to persons of older age at injury and/or with a history of pre-injury medical condition(s), to help them to obtain work and sustain employed for more years after injury.