Objective Minimally invasive surgical pulmonary vein isolation (SMI-PVI) is an emerging therapy for the treatment of symptomatic drug-refractory atrial fibrillation (AF). Nevertheless, the midterm and long-term results of SMI-PVI remain unknown. The aim of this retrospective multicenter study was to report on midterm efficacy and safety of SMI-PVI. Methods The study design was retrospective, multicentric, and observational. From July 2005 to November 2011, a total of 86 patients with drug-refractory paroxysmal or persistent AF underwent SMI-PVI in three centers. Patients were eligible for SMI-PVI if they had symptomatic, drug-refractory AF or after failed transcatheter pulmonary vein isolation. Success was defined as absence of AF on 24- or 96-hour Holter monitoring during follow-up, in the absence of antiarrhythmic drugs (AADs). Results The mean ± SD age was 54 ± 11 years, and 78% were men. The median AF duration was 30 months (range, 2–203); paroxysmal AF was present in 86% of the patients, persistent in 14%. Fifteen patients (17%) underwent previous transcatheter ablations. After a median follow-up of 24 months (range, 6–78), 72% of all patients were free from atrial arrhythmias without the use of AADs. With AADs, this was 83%. Major perioperative adverse events occurred in 7 patients (8%). Conclusions This retrospective multicenter study shows that SMI-PVI is effective at a median follow-up of 24 months for the treatment of mostly paroxysmal drug-refractory AF. Perioperative adverse events do remain a point of caution.