Background <p>Our original birth-size reference charts (the Olsen charts), published in 2010 (weight, length, head circumference) and 2015 (body mass index), had limitations, including manually defined exclusion criteria and the use of completed gestational weeks. This project aimed to update the charts and address those limitations.</p> Methods <p>We analyzed de-identified data from infants admitted to Pediatrix NICUs between 2010 and 2022. In contrast to prior work, newborns with mortality before discharge, multiple births, and congenital anomalies were included; only statistical outliers (defined as two times the interquartile range below the first quartile or above the third quartile) were excluded. Gestational age was recorded in days, and penalized b-splines were used as curve smoothers. Birth weight, length, head circumference, and BMI models were developed using a randomly selected subset and validated by the remaining data. Proportions of SGA and LGA infants were compared to the 2010 Olsen and 2013 Fenton charts.</p> Results <p>One million one hundred fifty-two thousand seven hundred twenty infants born between 22 and 41 weeks of gestation were included. No overfitting bias was observed in the new charts. These charts more accurately identified the expected 10% of newborns as SGA/LGA compared to existing charts.</p> Conclusions <p>The use of refined methods in new charts improved birth size referencing for NICU infants.</p>