SummaryBackground: Transcatheter device closure of atrial septal defects (ASD) is an alternative to surgery, but experience is limited in adults, especially in those with large (> 26 mm) defects.Hypothesis: We investigated the safety, efficacy, and learning curve for closure of ASD and patent foramen ovale (PFO) using the Amplatzer device.Methods: In all, 101 procedures were carried out in 100 consecutive adult patients in a single cardiac center between July 1998 and August 2002.Results: Preprocedure diagnosis was ASD and PFO in 50 patients each. A device was deployed in 94 of 101 attempts (93%) in 94 of 100 patients (94%). Atrial septal defect device sizes were 10-38 mm, median 24 mm, and 40% were > 26 mm. Major complications occurred in 2 of 100 patients (2%). One ASD device displaced requiring surgery within 24 h and one patient with PFO experienced pericardial tamponade; there were no deaths. Local vascular complications occurred in 4 of 100 (4%) and late complications in 4 of 100 (4%) patients. Patent foramen ovale closure was quicker (p < 0.001), required less radiation (p = 0.04), and was associated with fewer local vascular complications than ASD closure (p = 0.04). Deployment of ASD devices > 26 mm was not associated with increased complications, length of procedure, or radiation compared with devices ≤ 26 mm (all p > 0.05). Complications in the first 35 patients were more frequent than in subsequent patients: 7 of 35 (20%) versus 3 of 65 (4.6%) (p = 0.04); proce-