OBJECTIVES To estimate the cost-utility of sacral anterior root stimulation (SARS, using the Finetech-Brindley device) compared to medical treatment (anticholinergics + catheterization) in complete spinal cord injured patients with a neurological bladder. METHODS A probabilistic Markov model was elaborated with a 10-year time-horizon, one-year cycles and a 2.5% discount rate. Three irreversible states were defined: 1) treatment without urinary complication; 2) surgery for urinary complication (sphincterotomy, urinary derivation); 3) death. Reversible states (urinary calculi; Finetech-Brindley device failures) were integrated in the two first irreversible states. A systematic review and meta-analysis were performed to estimate transition probabilities and Quality Ajusted Life Years (QALYs). In the perspective of the French Healthcare System, costs were estimated from a published comparative cost-effectiveness research (Neurosurgery 2014;73:600), and through simulations using the 2013 French prospective payment system (PMSI) classification. RESULTS In the primary analysis, the cost-utility ratio was 10,647€/QALY gained. At a 30,000€ ceiling ratio the probability of SARS being cost-effective compared to medical treatment was 63%. If the French Healthcare System reimbursed SARS for 200 patients/year the two first years and 50 patients/year during 8 years (anticipated target population) the expected incremental net health benefit would be 222 QALYs, and the expected value of perfect information (EVPI) would be 4,570,000€. The highest partial EVPI is reached for transition probabilities toward urinary calculi (4,420,000€). With discount rates of 1% and 6% the cost-utility ratios were 6,951 and 19,770€/QALY gained, and the probabilities of SARS being cost-effective were 66% and 58%, respectively. CONCLUSIONS Our model shows that SARS using Finetech-Brindley device offers the most important benefit and should be considered cost-effective at a 30,000€ ceiling ratio. Despite a high uncertainty, EVPI and partial EVPI may indicate that further research would not be profitable to inform decision making.