Spinal Cord Injury (SCI) is damage to the spinal cord that results in loss of mobility and sensation below the level of injury. The incidence of SCI varies amongst countries. For example there are 12.7 in France and 59 in the United States of America, new cases per million populations each year. This type of disability makes most SCI individuals rely on a wheelchair for their mobility. They can traort themselves from one place to another using a manual wheelchair with the speed and energy expenditure similar to normal subjects. Using this wheelchair helps these patients' mobility but it still has some problems. The main problems are the restriction on mobility from architectural features in the landscape, and a number of health issues due to prolonged sitting. Osteoporosis, joint deformities, especially hip joint adduction contracture resulting from prolonged wheelchair use. SCI individuals often undergo various rehabilitation programs for walking and exercises. As expected, walking is a good exercise for paraplegics to maintain good health, decrease urinary tract infections and improve cardiovascular and digestive functions and psychological health. A variety of orthoses have been designed for paraplegic subjects to help them to stand and walk. However, patients experience some problems when they are using these orthosis and the most common one is high energy consumption during walking. Generally, researchers have used clinical evaluations (e.g. motion analysis) rather than mathematical modeling to investigate the effect of parameters like hip abduction or ankle flexion on energy consumption of paraplegic subjects during walking with orthosis. This study aimed to develop a mathematical model of both upper limb and lower limb to investigate the effects of various gait parameters on energy consumption of spinal cord injury subjects during walking with a mechanical orthosis. The skeletal model was developed based on the new type of reciprocal gait orthosis (RGO) degrees of freedom and it was assumed that the paraplegic subjects used elbow crutches to walk. Three paraplegic subjects with the lesion at level T12 were recruited in this research study. The kinematics and kinetics of the subjects were evaluated by Qualysis motion analyzer system and a Kistler force plate. These results were used to verify the validity of the mathematical model and to improve that. After verification of model by the results of clinical tests, the inverse dynamic of the model was solved and the energy expenditure per each walking cycle was calculated. Finally the effect of parameters including hip abduction, knee flexion, ankle flexion, heel height, shoulder abduction and crutch length on the energy consumption during walking were investigated by a parameter study. Hip abduction and knee flexion was changed from zero to 10 degrees and ankle flexion was studied from 15 degrees dorsiflexion to 15 degrees plantar flexion. Heel height varied from zero to maximum of 7 centimeters. Crutch length and shoulder abduction were chosen accordingly. All increment steps were0.1 degree for angles and 1 centimeter for lengths. In each study, one parameter was varied in its range while the other were fixed. Five different sets of parameters in each study were considered. Moreover, the influence of physiological parameters including body weight and length on the optimum value of each parameter were investigated. Key words: Modeling, Reciprocal Gait Orthosis, Walking, Energy Consumption, spinal cord injury.