Abstract: This paper describes a new method for determination of the mass concrete measuring points for inversion calculation of thermal parameters using the principle of the highest sensitivity to solve the problem that the inversion accuracy of thermal parameters depends on the selected locations of measuring points and the stability of optimization algorithm. This method uses a heat-fluid coupling algorithm to simulate the temperature field in an under construction concrete mass and a sensitivity analysis of the on-site observed temperature to determine the most sensitive points. It includes a new technique of gradually narrowing the search space and an inversion algorithm that improves the efficiency and stability of the genetic algorithm. Results show that the method can improve the efficiency, accuracy and stability of inversion calculations.