We wished to determine whether surgery in adults with difficult to control temporal lobe epilepsy (TLE) is more effective and less costly than medical therapy. We evaluated the costs and consequences of surgical versus medical treatment of TLE, using decision-analysis modeling and an intention-to-treat approach. Data on outcome probabilities and resource utilization were obtained from critical appraisal of the literature, local experience, a panel of experts, and community patient survey. Surgically treating 100 patients for 35 years results in 57 seizure-free patients and a cost of $8,117,911. Medically treating 100 patients results in 12 seizure-free patients and a cost of $10,741,425. Surgery required a larger initial expenditure than medical treatment ($2,775,640 vs. $856,970), but cost-time curves intersect at 8.5 years and surgery remained cheaper thereafter. Results were unchanged by extensive sensitivity analyses. The major determinant of surgical dominance is its greater effectiveness (seizure-free rate) as compared with medical treatment. Surgery remained less costly, with concurrent effectiveness rates >41 and <30%, respectively, for surgical and medical treatment. Available evidence indicates that TLE surgery is more effective and cheaper than medical treatment, a win-win situation. However, adequately designed, controlled trials are necessary to determine more accurately and comprehensively the effectiveness of each therapeutic alternative.