Although multibody paediatric pedestrian human models are widely used to study pedestrian crashes, the biofidelity of these models has only been evaluated for limited loading conditions. The current study aims to evaluate the biofidelity of multibody paediatric models developed by scaling a baseline 50th percentile adult male model. The biofidelity of the baseline adult model was thoroughly evaluated prior to developing the scaled paediatric models. Next, three‐ year old, six‐year old, and 12‐year‐old pedestrian human models were developed using a structure‐ based scaling method. Paediatric experimental data, including component‐level tests on head and neck and blunt impact tests for thorax, abdomen and pelvis, were collected from the literature and used as reference responses. Simulations were performed using the paediatric models and the model responses were compared to the reference responses using an objective rating method. While the baseline adult model showed a “good” quality score in various loading conditions, the quality score of the scaled paediatric pedestrian models was rated as “acceptable”. The twelve‐year old model showed the best biofidelity of the three child models. This paper demonstrates the quality of the multibody paediatric human models that can be developed using only a scaling technique.