Roberton’s developmental sequences for components of the overarm throw for force (Roberton & Halverson, 1984) is the most valid and reliable developmental tool used to assess overarm throwing in children. However, there is very little biomechanical data on children’s throwing motions and on the actual quantitative differences within developmental levels. Therefore, the purpose of this investigation was to assess overarm throwing motions in children and compare and contrast the results of a developmental component analysis against biomechanical data collected on the same children. Twenty-six children (11 boys, 15 girls), ages 3-12 participated in the study which was conducted at the American Sports Medicine Institute in Birmingham, AL. Children were tested on five maximal effort throws. Four developmental components (stepping action, trunk action, humerus action, and forearm action) were analyzed using data collected from two 60 Hz video cameras set at perpendicular angles to the thrower. The same throws were also analyzed using three-dimensional high-speed motion analysis (240 Hz) with 19 kinematic variables quantitatively describing differences in levels of the four components. Results indicated that, in general, the quantitative biomechanical data agreed with the component analysis descriptions of body segment movements. Differences between biomechanical data and component descriptions of movement were noted in level one trunk action of the component analysis (defined as having no rotation), where some rotation of the pelvis and upper torso was present (average pelvis angular velocity = 215 ± 74º/s, average upper torso angular velocity = 426 ± 179º/s). One other difference between the biomechanical data and component analysis was within level one forearm action. Level one forearm action suggested that there would be no increase in external rotation (no lag) as the child rotates forward. The biomechanical data suggested that there was increasing external rotation throughout the arm acceleration phase of the throw in the level one forearm throws (40 ± 23º/s). Comparisons of the component analysis and biomechanical data within the highest level of trunk action, humerus action and forearm action were limited due the limited number of throwers that exhibited the highest level of these components. Keyword(s): research