The inclusion of information from biomechanical research on human movement has led to many advances in how coaches, teachers, and parents teach fundamental and advanced motor skills. However, the development of overhand throwing in children has not been able to utilize biomechanical research at the most fundamental level of teaching. Developmental models have been utilized to assist in developing teaching methodology but have been limited to defining throwing movement sequences from qualitative data analysis. The integration of biomechanical research and motor development research would significantly impact traditional views on teaching overhand throwing. Therefore, the purpose of the current investigation was to compare the effectiveness of two throwing instruction protocols on motorically delayed children. One protocol utilized combined research from the areas of biomechanics and motor development (B/D) (N=11). The second protocol was based on task instructions developed by Graham et al. (1987) (N=9). Throwing levels were evaluated using a component approach (Roberton & Halverson, 1984). Instruction and practice time was limited to 20 minutes per day for four sessions during an after school motor skill intervention program. An independent t-test was implemented to determine group differences in improvement in developmental levels. Results did not show a statistical difference in improvement between groups (p=.065). However a trend towards overall improvement could be seen more in the B/D group as the mean improvement across all components for the B/D instruction group was almost double compared to the Graham instruction group (3.00 – 1.55 levels). Specific group component analysis revealed that the majority of change was evident in the stepping action and forearm action. Subjects in the B/D group increased an average of 1.73 levels (component level scale: 1-4) in the stepping action as compared to an average increase in the Graham subject group of .9 levels. Also, six subjects in the B/D group improved forearm action one level where only one subject in the Graham group had improved forearm action. These results seem promising considering the limited sample size and limited instruction time available for each protocol. Further research on the integration of biomechanical information in motor development research should focus on allotting more time for instruction and practice as well as comparing the B/D protocol with other throwing protocols. Keyword(s): elementary education, physical activity, research