We reviewed pre-closure echocardiograms on all patients undergoing transcatheter atrial septal defect (ASD) closure with the Bard double-umbrella occluder device aided by simultaneous transesophageal echocardiography to determine precatheterization predictors of outcome. Transesophageal echocardiograms were performed on 28 of 132 patients (22%) undergoing device closure (age = 3-72 years, mean = 14 years; weight = 15-68 kg, mean = 35 kg). Three devices were removed because of unstable position. Of the remaining 25 patients, 21 had effective closure (residual flow diameter < or = 3 mm) and 18 had favorable arm position (device arm on proper side of the septum and not in contact with an atrioventricular valve leaflet). Only ASD size predicted effective closure. All patients with a maximum defect size of < 13 mm had effective closure. Among the 17 patients with defects > or = 13 mm, 10 had effective closure, 4 had significant residual flow, and 3 had devices removed for unstable position. Atrial dimensions and rim size did not predict effective closure. There were no pre-closure predictors of favorable arm position which was associated only with the size of the device implanted.