Device closure of atrial septal defect (ASD) is commonly performed in older children and adults. Infants and toddlers (age <4 years) are seldom referred for ASD closure due to size constraints. However, in many cases device ASD closure can be performed in this population. Between 2002 and 2012, 61 infants and toddlers were taken to the catheterization laboratory at our institution for ASD closure. Precatheterization transthoracic echocardiograms, intracatheterization transesophageal echocardiograms, and catheterization reports were reviewed. Fifty-three infants and toddlers presented for percutaneous ASD occlusion. Forty-eight (79 %) underwent successful closure, and 13 were referred for surgery without device attempt (n = 8) or after unsuccessful device occlusion (n = 4). Median age and weight at time of ASD closure were 2.99 years (range 0.3-3.8) and 11.7 kg (range 3.7-16.5). The device-to-septal length ratio was 0.81 (range 0.44-1.03). The 12 unsuccessful cases occurred in patients with larger defects (ASD diameter 17.5 ± 6.1 vs. 12.1 ± 4.2, p < 0.01). Deficient rims (absent or ≤ 4 mm) were seen in 9 of 12 (75 %) unsuccessful cases and in 19 of 41 (46 %) successful cases (p = 0.12). Multivariate analysis showed that patient size and ASD size were not independently associated with procedural success but that ASD size-to-patient weight ratio <1.2 (hazard ratio 9.5 [range1.7-17]) was associated with successful ASD closure. ASD device occlusion can be safely achieved in small children. An ASD size-to-patient weight ratio >1.2, not absolute patient weight or age, is associated with failure of the percutaneous approach. The midterm outcomes in these young patients are excellent.