Background: Clinicians have observed preterm infants in the neonatal intensive care unit growing disproportionally; however, the only growth charts that have been available were from preterm infants born in the 1950s which utilized the ponderal index. Prior to creating the recently published BMI curves, we found only 1 reference justifying the use of the ponderal index. Objectives: To determine the best measure of body proportionality for assessing growth in US preterm infants. Methods: Using a dataset of 391,681 infants, we determined the body proportionality measure that was most correlated with weight and least correlated with length. We examined the sex-specific overall correlations and then stratified further by gestational age (GA). We then plotted the body proportionality measures versus length to visualize apparent discrepancies in the appropriate measure. Results: The overall correlations showed weight/length3 (ponderal index) was the best measure but stratification by GA indicated that BMI (weight/length2) was the best measure. This seeming inconsistency was due to negative correlations between ponderal index and length at each GA. BMI, on the other hand, had a correlation with length across GAs, but was uncorrelated with length within GAs. Both ponderal index and BMI were positively correlated with weight. Conclusions: BMI is the appropriate measure of body proportionality for preterm infants, contrary to current practice.