We retrospectively identified 92 children aged </=9 years with isolated secundum atrial septal defect (ASD) with preoperative height or weight z score < or =1.0 (16th percentile), respectively. Children with additional cardiac malformations, chromosomal abnormalities, recognized syndromes, pulmonary hypertension, heart failure, prematurity (<34 weeks' gestation), or use of cardiac medications were excluded from further analysis. After exclusion, 24 low-weight and 25 low-height children were compared with 48 and 50 age-, gender-, and size-matched normal controls, respectively. Univariate and multivariate analyses were performed separately for height and weight groups for achieving an end point of improvement in height or weight z score by +0.5 after repair. There were no significant preoperative demographic differences. After repair, 50% of low-weight patients improved +0.5 SD in weight within 2.6 years versus 5.6 years for controls (p = 0.02). Similarly, 50% of low-height patients improved +0.5 SD in height within 1.7 years versus 11.6 years for controls (p <0.001). Multivariate analysis also indicated a benefit of repair with relative risks of 2.35 (95% confidence interval 1.20 to 4.62; p = 0.01) and 5.11 (95% confidence interval 2.45 to 10.65; p <0.0001), respectively, for improving height or weight by +0.5 SD in patients with ASD relative to controls. Furthermore, children heavier at repair were less likely to improve in weight (relative risk 0.56 [0.34, 0.92] per 1.0 z [p = 0.02]). Likewise, children older at repair were less likely to improve in height (relative risk 0.77 [0.64, 0.92] per year of age [p = 0.003]). We conclude that undersized children with ASD exhibit improved growth after repair. Poor growth should be considered a relative indication for early ASD repair.