In resource-constrained facilities or during resuscitation, immediate paediatric weight estimation remains a fundamental challenge. We aimed to develop and validate weight estimation models based on ulna length and forearm width and circumference measured by simple and portable tools; and to compare them against previous methods (advanced paediatric life support (APLS), Theron and Traub–Johnson formulas).

Cross-sectional analysis of anthropometric measurements. Four ulna- and forearm-based weight estimation models were developed in the training set (n 1016). Assessment of bias, precision and accuracy was examined in the validation set (n 457).

National Children’s Study-Formative Research in Anthropometry (2011–2012).

Multi-racial/ethnic infants and children aged <6 years (n 1473).

Developed Models 1–4 had high predictive precision (R2=0·91–0·97). Mean percentage errors between predicted and measured weight were significantly smaller across the developed models (0·1–0·7 %) v. the APLS, Theron and Traub–Johnson formulas (−1·7, 9·2 and −4·9 %, respectively). Root-mean-squared percentage error was overall smaller among Models 1–4 v. the three existing methods (range=7·5–8·7 v. 9·8–13·3 %). Further, Models 1–4 were within 10 and 20 % of actual weight in 72–87 and 95–99 % of the weight estimations, respectively, which outperformed any of the three existing methods.

Ulna length, forearm width and forearm circumference by simple and portable tools could serve as valid and reliable surrogate measures of weight among infants and children aged <6 years with improved precision over the existing age- or length-based methods. Further validation of these models in physically impaired or non-ambulatory children is warranted.