Résumé : ABSTRACT: BACKGROUND: Children with growth retardation or short stature generally present with lower strength than children of the same chronological age. The aim of the study was to establish if strength was dependent on variables related to stature in a population of healthy children and to propose practical predictive models for the muscle functions tested. A secondary aim was to test for any learning effects concerning strength measured at two successive visits by children. METHODS: Hand grip, elbow flexion and extension, and knee flexion and extension were measured by fixed dynamometry in 96 healthy subjects (47 girls and 49 boys, aged from 5 to 17 years). RESULTS: For the present paediatric population, muscle strength was highly dependent on height. Predictive models are proposed for the muscle functions tested. No learning effect between the first and the second visit was detected for any of the muscle functions tested. CONCLUSIONS: This work shows that strength measurements using fixed dynamometry are reliable in children when using appropriate standardization of operating procedures. It underlines the particular relationship between body stature and muscle strength. Predictive equations may help with assessing the neuromuscular involvement in children suffering from various disorders, particularly those affecting their stature.