Upper extremity (UE) joint kinetics during aided ambulation is an area of research that is not well characterized in the current literature. Biped UE joints are not anatomically designed to be weight bearing, therefore it is important to quantify UE kinetics during assisted gait. This will help to better understand the biomechanical implications of UE weight bearing, and enable physicians to prescribe more effective methods for treatment and therapy, perhaps minimizing excessive loads and torques. To address this challenge, an UE model that incorporates both kinematics and kinetics has been developed for use with walkers instrumented with load cells. In this study, the UE joint kinetics are calculated for 10 children with cerebral palsy using both anterior and posterior walkers. Three-dimensional joint reaction forces and moments are fully characterized for the wrist, elbow, and shoulder (glenohumeral) joints for both walker types. Statistical analysis methods are used to quantify the differences in forces or moments between the two walker types. Comparisons showed no significant differences in kinetic joint parameters between walker types. Results from a power analysis of the current data are provided which may be useful for planning longer term clinical studies. If risk factors for UE joint pathology can be identified early, perhaps a change in gait training routine, walker prescription, or walker design could prevent further harm.