The invention relates to a transaction object selection optimization method for electric energy dealers based on a conditional value at risk. The method includes: classifying generating producers andusers based on the capacity according to the generating capacity of the generating producers and the power consumption capacity of the users, and obtaining distribution characteristics of the generating producers and the electricity prices of the users in each capacity level according to historical transaction data; extracting samples of the generating producers and the electricity prices of the users in each capacity level by employing a Monte-Carlo method; calculating a loss function of the dealers and the generating producers and the users in each capacity level during transaction; calculating the conditional value at risk during the transaction of the dealers; establishing a transaction object selection optimization model by regarding the minimum conditional value at risk as the optimization target; and performing model solution by employing a particle swarm optimization method. According to the method, risks during transaction between the electric energy dealers and objects of different capacities in terms of electricity purchase and electricity sales are comprehensively considered, the transaction objects are selected in an optimized manner, bases are provided for making of an electric energy transaction policy for the electric energy dealers, and important theoretical and practical values are achieved for promotion of marketization reform of the electricity sales side.