Load demand forecasting is a key point in electric power system operation and planning. It is used to determine the capacity of generation, transmission and distribution system. In Ethiopian electric power system, the electric demand has been more rapidly increasing than generation expansion process. Due to dominant energy consumption of firewood and charcoal, environmental degradation and deforestation is continuously increased. To minimize those problems, multi-objective model preceded by electric demand forecasting was developed in this work. Main task which are considered in this thesis includes; investment and generation cost of power generating units, and assessment of environmental impact and carbon-trading benefit. Long term load forecasting for the whole country was carried out using Artificial Neural Network (ANN) and based on long term demand forecasting, optimal power generation expansion planning is done using suitable written code in MATLAB for dual-simplex method for each developed model. Among 31 candidate power plants, only power plants with least investment unit cost that will put into operation are selected by the thesis. Accordingly, five candidate power plants namely, Mendaya, Tams, Bako Abo, Genale Dawa III and Border are selected for next ten years through optimal power generation expansion process. Key words: Carbon-Trade, Long-Term Load Forecasting, Optimal Power Generation Expansion Planning