Purpose <p>Body mass index (BMI), is a nutritional index based on standing height and weight. Children with early onset scoliosis (EOS) have shorter height due to their spine curvature. In these patients’ pulmonary function tests, which are also based off height, arm span is standardly used as a surrogate for height to allow for normative comparisons. Arm span has not been used for nutritional assessments in children with EOS.</p> Methods <p>We used both arm span and measured height to calculate BMI, and BMI as a percent of predicted published norms (BMI%ile) to assess the error produced using height in this population of children. We used a Wilcoxon signed-rank test to compare standard height BMI (BMI) vs arm span BMI (asBMI) and BMI%iles, Spearman correlation to assess the influence of age or coronal curve angle on any differences, and an adjusted linear model for the coronal curve’s impact on BMI.</p> Results <p>BMI and BMI%ile consistently over-estimated nutritional status compared to asBMI with substitution of arm span for height. 15.6% of patients had a BMI%ile of &lt; 1% using height compared to 44.4% using arm span. The difference between BMI and asBMI correlated with the coronal curve magnitude but not with age. The adjusted linear model showed that for each degree of Cobb angle, the difference between BMI and asBMI increased by 0.0294&#xa0;kg/m² (<i>p</i> = 0.0062).</p> Conclusion <p>Arm span BMI reveals lower, more accurate nutritional status in children with EOS than standard height BMI.</p> Level of evidence <p>IV.</p>