Background: Atrioventricular block (AVB) was reported following atrial septal defect (ASD) occlusion by Amplatzer septal occluder (ASO). The main hypothesis is the impingement of ASO on the conduction system. However, in knock out mice, hypoplasia of the conduction system was described in association with NKX 2.5, Id2 and GATA4 mutations. We hypothesized that mutations of these genes predispose to AVB following ASO in humans. Methods: Patients with new onset AVB or worsened AV conduction following ASO were compared to controls matched for previously identified risk factors (device size, size/height ratio). Clinical evaluation, three-generation family history, echocardiograms, electrocardiograms and genetic sequencing of candidate genes were performed. Results: Twenty eight patients (6.4±4.2 yo, 23.2±14.9 kg, ASO size 25.4±5.2 mm) and 28 controls were enrolled. Second or third degree AVB occurred in 15 patients, first degree AVB (AVB-1) in 13. AVB resolved in 24/28 patients (median 8 days, 0.5-240). AVB-1 remains in 4, 4.17±2.35y later. A positive family history of cardiac malformation (n=17/55) did not correlate with AVB after ASO (OR= 1.6, CI 0.5-5, p=0.3). Gene sequencing identified a mutation affecting the NKX 2.5 homeodomain (Arg189Leu) in a patient with preexisting borderline PR evolving towards AVB-1/2. No coding mutations were detected in GATA4 or ID2. Conclusion: In this series, mutations in NKX2.5, GATA4 or ID2 were rare events, and mechanical stress of the ASO on the conduction system remains a plausible cause for AVB. However, this study provides the first proof-of-principle that NKX2.5 homeodomain mutations can predispose to worsening of AVB after ASO.