Background <p>There are expert-agreed upon malnutrition indicators for infants, but the meaningfulness of these indicators in hospitalized neonates is unknown.</p> Methods <p>Eighty-four term infants hospitalized in a level IV neonatal intensive care unit had body composition measurements and other anthropometric data extracted from the medical record. Linear regression using maximum likelihood estimation was performed to evaluate the relationship of malnutrition status (assessed at time of body composition assessment) with each body composition metric (body fat percent z-score, fat mass z-score, fat-free mass z-score) and with toddler body mass index (BMI) z-score. Linear regression was also performed to evaluate the relationship between each body composition metric and toddler BMI z-score.</p> Results <p>There was a statistically significant negative association between neonatal malnutrition diagnosis with a -0.83 change in fat-free mass z-score (95% CI -1.61 to -0.05, <i>p</i> = 0.04). There was no statistically significant association between body fat percent or fat mass and neonatal malnutrition diagnosis. There was no statistically significant relationship between each infant body composition metric and toddler BMI z-score in unadjusted linear regression models.</p> Conclusions <p>Neonatal malnutrition diagnosis is associated with lower fat-free mass in critically ill infants with a variety of conditions.</p>