Surgical closure of atrial septal defect (ASD) provides excellent results. Given the increasing popularity of percutaneous techniques, a comparison between the 2 methods is needed. Between December 1988 and June 2003, we performed 1284 procedures in 1268 consecutive patients with isolated secundum ASD. Five hundred and thirty-three patients underwent surgical repair of ostium secundum ASD (group A). Seven hundred and fifty-one consecutive patients underwent percutaneous ASD closure (group B). The following outcomes were studied: mortality, morbidity, hospital stay, and efficacy. There were no postoperative deaths. The overall rate of complications was higher in group A than in group B: 44% (95% CI 39.8%-48.2%) versus 6.9 % (95% CI 5%-8.7%) (P < .0001). Major complications were also more frequent in group A: 16% (95% CI 13%-19%) versus 3.6% (95% CI 2.2%-5.0%) (P = .002). Multiple logistic regression analysis showed that surgery was independently strongly related to the occurrence of total complication (odds ratio [OR] 8.13, 95% CI 5.75-12.20) and of major complications (OR 4.03, 95% CI 2.38-7.35). The occurrence of minor complications was independently related to surgery (OR 7.33, 95% CI 4.75-11.02), childhood (OR 1.52, 95% CI 1.01-2.34), and presence of systemic hypertension (OR 1.35, 95% CI 1.01-4.41). Hospital stay was shorter in group B (3.2 +/- 0.9 vs 8.0 +/- 2.8 days, P < .0001). Percutaneous ASD closure provides, in experienced hands and in highly specialized centers, excellent results with a lower complication rate and requires a shorter stay in hospital.