OBJECTIVES: Catheter ablation is increasingly used to maintain sinus rhythm in atrial fibrillation (AF) patients unresponsive to antiarrhythmic drugs (AADs). We compared medical costs in Medicare-aged AF patients following successful vs unsuccessful ablation. METHODS: In this retrospective study, AF pts with 1) an index ablation; 2) ≥12 months’ medical/pharmacy coverage pre- and post-index; 3) ≥2 AF inpatient/outpatient visits within 6 months and AAD treatment within 12 months of index ablation were identified from the MarketScan® Medicare database (January 2002-June 2007). Ablation success was defined as absence of AAD treatment 6-12 months post ablation. RESULTS: A total of 135 AF patients (67% men, mean 73 yrs) were included; ablation was successful in 69 and failed in 66 patients. Most patients (97% with successful vs 94% with failed ablation) underwent only 1 ablation procedure during the 12-month study. After successful ablation, patients discontinued AAD in (mean) 54 days. Use of rate-control and anticoagulant drugs declined after successful ablation (67% vs. 87% and 64% vs. 86% patients, respectively), but remained largely undiminished after failed ablation (70% vs. 74% and 82% vs. 88% patients, respectively). Mean (median) per-patient costs per ablation were $13,655 ($11,795) for successful vs. $17,294 ($11,778) for failed ablation. Other AF-related costs over 1 yr post index ablation were $2394 ($7677) for successful vs. $2703 ($4478) for failed ablation patients. Overall annual per-patient costs were lower in patients with successful (mean $16,049; median $17,135) vs. failed ($19,997; $26,635) ablation (P =0.07). CONCLUSIONS: Ablation failed in half of ‘real-world’ Medicare-aged AF patients, and few underwent repeat ablation. Overall costs were higher for failed vs successful ablation patients, possibly because of differences in AF-related issues, complications, and ablation methods. Over time, this cost differential would likely increase if failed ablation patients underwent repeat procedures. Identification of predictors for ablation success may reduce medical costs.