Study design: Prospective, multicentred follow-up (FU) observational study. Objectives: Prospectively evaluate survival, complications, re-admissions and maintenance of clinical outcome in people experiencing traumatic spinal cord injury (SCI). Setting: Seven spinal units and 17 rehabilitation centres participating in the previous GISEM (ie Italian Group for the Epidemiological Study of Spinal Cord Injuries) study. Method: A total of 511 persons with SCI, discharged between 1997 and 1999 after their first hospitalisation, were enrolled. A standardised questionnaire was administered via telephone. Results: Of the 608 persons originally enrolled, 36 died between discharge and follow-up (mean 3.870.64 years). Of the remainder, 403 completed telephone interviews, 72 refused to participate and 97 could not be contacted. More than half of the patients interviewed (53.6%) experienced at least one SCI-related clinical problem in the 6 months preceding interview; the most frequent being urological complications (53.7%). At least one re-admission was recorded in 56.8% of patients between discharge and FU interview. Of the patients interviewed, 70.5% reported bowel autonomy and 86% bladder management autonomy. On multivariate analysis, lack of bowel/bladder autonomy was the most common variable with a strong predicting value for mortality, occurrence of complications and re-admissions. Conclusion: Re-admission and major complications seem common after SCI and should be considered when planning facilities. Failure to obtain bowel/bladder autonomy upon discharge from rehabilitation proved to be the most common predictive factor of poor outcome during the period between discharge and FU interview.