PURPOSE: To compare resective surgery and medical therapy in a cost-effectiveness analysis in a multicenter cohort of adult patients with partial intractable epilepsy. POPULATION AND METHODS: Adult patients with partial, medically intractable, potentially operable epilepsy were eligible and followed every year over five years. Effectiveness was defined as one year without seizure. The long-term costs and effectiveness were extrapolated over the patients' lifetime with a Markov model. Productivity (indirect costs) and quality of life (QOLIE-31, SEALS) were also assessed. Changes before and after surgery were compared between the two groups. RESULTS: Two hundred and eighty-nine patients were included (119 with surgery, 161 medically treated, six not eligible, three lost to follow-up). One year after surgery, 81% of the patients were seizure-free; at two and three years, this rate was 78%. In the medical group, these rates were 10, 18, and 15%, respectively. The cost of the explorations was euro 8464; including surgery, it was euro 19,700. In the medical group, the average annual direct costs were between 3500 and euro 6000. At two years after surgery, the annual direct cost decreased to euro 2768, at three years, it was euro 1233, predominately antiepileptic drug costs. Surgery became cost-effective between seven and eight years. In the surgical group, all the quality-of-life scores improved at one year after surgery and were stable during the second and third years. CONCLUSION: Surgical therapy was cost-effective at the middle term even though indirect costs were not considered.