Coal consumption curve of the thermal power plant is usually obtained from the performance parameters which are provided by the manufacturer or from the thermal test data. These curves remain unchanged for a long time and are incompatible with the actual operation situation of the unit. Therefore, a method of coal consumption curve fitting of the thermal power plant units based on genetic algorithm is proposed. The quadratic function is used as the objective function; appropriate parameters such as initial population size, crossover rate and mutation rate are set; the unit’s actual coal consumption curves are fitted. The fitting curve of the proposed method is compared with that of the least squares method. The results indicate that fitting effect of the former is better than that of the latter. It is indicated that the proposed method can best approximate the known data in the curve fitting, and they can real-timely reflect the interdependence between unit generation and coal consumption.