The invention relates to the technical field of electric power, in particular to a Gini coefficient method for making a monthly electric power trade schedule of a directly dispatched thermal power generation unit. A Gini coefficient is used for judging set electric quantity fairness, and a set electric quantity fairness constraint condition is met when a set residual load rate meets a requirement that the Gini coefficient is smaller than 0.3. In the method, the residual load rates of all thermal power generation units to be subjected to plan making meet a Gini coefficient constraint, and the comprehensive cost of the thermal power generation unit is lowest when constraint conditions including the set electric quantity fairness, energy saving, environment protection and the like are guaranteed. The set residual load rate is simultaneously constrained by a load deviation threshold value and the Gini coefficient, the range of the load rate deviation threshold value can be properly broadened on a premise that the residual load rate meets the Gini coefficient constrain, and therefore, the comprehensive cost of the thermal power generation unit is lowered. The comprehensive cost of the monthly electric power trade schedule of the directly dispatched thermal power generation unit on the basis of a Gini coefficient constraint method is lower than the comprehensive cost of an existing load rate deviation method and an existing comprehensive cost weighting method.