Study design: Case series of a consecutive sample. Retrospective audit. Objective: To analyze the long-term safety and efficacy of intrathecal baclofen (ITB), and technical incidents. Setting: Neurosurgical and Physical Medicine Departments of two university hospitals in western France. Methods: The medical records of 40 patients who underwent ITB pump placement for the treatment of severe chronic spasticity were reviewed. Patients were eligible independently of the origin of the spasticity (spinal cord origin 33, brain damage 8). They underwent a final assessment with clinical examination and questionnaire in 2001. Ashworth scale scores were assessed, patient satisfaction was rated on a visual analog scale (VAS), functional independence before and after treatment was classified as bed-ridden, wheelchair dependant or ambulant, and the frequency and nature of complications were noted. Results: The average follow-up period was of 4 years. The average Ashworth score at the final assessment was 1.870.6. Average patients satisfaction was 7.4/1072.21 on VAS. In all, 85% would have undergone the procedure again if they had to make the decision. In 85% of the cases the ambulation status was unchanged. Technical incidents occured at least once in 37% of the patients (due to the catheter in 58% and to the pump in 42%). They included catheter disconnections (4), migration (4), kinks (3), obstruction (3), development of fibrosis (3), disconnection of pump reservoir (2), porosity of pump membrane (2), unexplained pump dysfunction (4) leakage, and subcutaneous collection (5). Severe pharmacological side effects requiring transfer to intensive care unit occured in 12% of cases, 80% of which were directly related to pump refill procedures. Conclusion: ITB remained effective in the long term and patients were satisfied. Nevertheless, complications were frequent, involving mainly the catheters, which would require further technical improvements.