This study reports on experience with transcatheter closure of congenital ventricular septal defects (VSDs) with Amplatzer septal occluders. From January 2000 to April 2005, transcatheter Amplatzer device implantation was attempted in 122 patients with congenital VSD (30 with muscular, 87 with perimembranous, and 5 with residual postsurgical repair of conotruncal malformations). Patient mean age was 15 years (range, 6 months to 64 years), and mean weight was 35 kg (range, 5.8 to 102 kg). The VSD mean size was 7 mm (range, 4 to 16 mm), mean Qp/Qs was 2.1 (range, 1.3 to 4), and mean fluoroscopy time was 32 minutes (range, 5 to 129 minutes). All procedures were performed with the patient under general anesthesia and guided by fluoroscopy and transesophageal echocardiography. The device size chosen was usually 1- to 2-mm larger than the maximum defect size as assessed by either the echocardiographic or angiographic views that were judged most reliable. Amplatzer muscular devices were placed in 47 patients, and the membranous devices were placed in 72 patients. The procedure was not performed in 3 patients with perimembranous VSD because of the impossibility of achieving an adequate long sheath position in 1 patient, onset of complete atrioventricular (AV) block during catheter manipulation in 1 patient, and the presence of aortic valve prolapse preventing a safe device placement in 1 patient. Satisfactory device implantation was achieved in 119 of 122 patients (97.5%): a tiny smoke-like residual flow through the device was often seen immediately after the procedure (50%); residual shunting was detectable in 19% after 24 hours and in only 4% at 6 months. The following additional catheter interventions were performed simultaneously: balloon pulmonary valvuloplasty in 3 patients, device closure of atrial septal defects in 2 patients, coil occlusion of the arterial duct in 1 patient, stenting coarctation in 1 patient, and stenting of the right pulmonary artery in 1 patient. Minimal aortic regurgitation developed in 3 patients, and minimal tricuspid regurgitation in 3 patients; no patient required additional treatment. Device embolization occurred in 3 patients (1 patient with muscular VSD, 2 with perimembranous VSD); catheter retrieval and implantation of a second device was successfully performed in all patients. Transient left bundle branch block occurred in 2 patients, and transient first-degree AV block in 1 patient. Among the perimembranous VSD cases, complete AV block occurred acutely (within 48 hours) in 3 patients, requiring a pacemaker in 1 patient; complete heart block occurred in the other 2 patients after 5 and 12 months, requiring pacemakers. There was no mortality. Transcatheter closure of muscular and perimembranous VSDs offers encouraging results: 96% complete closure at midterm follow-up. Complications are limited; the most relevant appears to be device-related complete heart block in perimembranous VSD. Greater experience and long-term follow-up are required to assess the safety and effectiveness of this procedure as an alternative to conventional surgery.