Trigeminal neuralgia is a recurrent severe shooting neuropathic pain which can be managed both pharmacologically and surgically. However, there are no prospective data that compare these two therapeutic strategies. This study therefore aimed to assess long-term outcome in patients with intractable trigeminal neuralgia treated with oxcarbazepine and later with surgery. Fifteen patients (11 females) with trigeminal neuralgia intractable to available drugs (carbamazepine, phenytoin and baclofen), were prospectively followed for 13 years (1986–1999) with a total follow up time from onset of disease of 16±6 years (mean±SD), range 8–30 years. All patients were contacted in 1999 and 12 replied, two had died and one had last replied in 1996. Patients were first treated with oxcarbazepine 1200±600 mg daily dosage (mean±SD) and subsequently with surgery of their choice. The outcome measures used were: McGill Pain Questionnaire, Hospital Anxiety and Depression Scale, patient satisfaction questionnaire and clinicians' global evaluation. Pain control was initially achieved in all patients and oxcarbazepine was used continuously or intermittently for 4.0±3 years (mean±SD). Thirteen patients experienced some mild side effects and a dose-dependent hyponatraemia was noted. Subsequently, 12 patients required surgery (five microvascular decompressions and seven surgery at the level of the Gasserian ganglion) to control their pain and were followed up for 4.3±1.7 years post surgery (mean±SD). Three patients required repeat surgery to control their pain, which was successful in two. A further two patients continued with low dose medication post initially successful surgery. Three patients reported numbness and one hearing loss after surgery. Kaplan Meier analysis 3 years after oxcarbazepine use and then 3 years after surgery showed that the mean time for recurrence of pain after oxcarbazepine treatment was 10 months whilst for surgery it was 28 months ( P<0.0001). Pain free periods and types of complications post surgery varied and depended on the type of surgery performed. Due to the small numbers, it was not possible to analyse the different types of surgical procedures individually. Outcomes after any type of surgery were better on all evaluations and eight patients felt that they should have had surgery earlier. Oxcarbazepine is a potent antineuralgic drug with very good acceptability and tolerability. However, its effectiveness was rather short term necessitating surgical intervention. As surgery was associated with better outcome, patients may therefore benefit from having surgery earlier rather than later in the disease process in order to improve quality of life, freedom from medication and the need for regular follow up. Surgery does not provide pain relief for all patients. This is the first study that has compared outcome in a group of patients who have had both pharmacological and surgical treatments. As these data cannot be extrapolated to other antineuralgic drugs, similar comparative studies would be appropriate.