Reliable forecasting of the service life of building materials and products allows you to lay down the costs of repair work in a timely manner, which in modern economic realities is undoubtedly an urgent task. This paper presents the results of a study on the development and comparison with existing methods for determining the thermal fluctuation constants of the generalized Zhurkov equation. A new method is proposed for determining the thermal fluctuation constants of the generalized Zhurkov equation. Practical application of the methodology will make it possible to reliably predict the service life of building materials. The main goal is to develop a method for determining the thermal fluctuation constants of the generalized Zhurkov equation, characterized by higher reliability by reducing the number of operations entailing errors, while increasing the number of experiments conducted under identical conditions (increasing the sample when determining durability under constant operating conditions). To achieve this goal, it is necessary to solve a number of tasks: 1) analyze the main provisions of the thermal fluctuation concept; 2) develop a method for determining the thermal fluctuation constants; 3) to conduct a comparative analysis of the obtained results of determining the thermal fluctuation constants. The object of the study is the constants of thermal fluctuation. The subject of the study is a new method for determining thermal fluctuation constants. The main methods of scientific knowledge used in the development of the methodology are hypothetical (the hypothesis of a linear dependence of the change slope of direct temperatures in the coordinates of the logarithm of durability - stress) and experiment (determination of durability of samples under transverse bending under specified operating conditions). A new method was developed for determining the thermal fluctuation constants of the generalized Zhurkov equation. It allows you to determine constants by plotting only one straight line temperature and one control point at a different temperature. Application of the proposed technique allows increasing the number of samples tested in identical conditions while reducing labor costs for experimental research. An increase in the sample leads to an increase in the accuracy and reliability of predicting the service life of building materials.