Abstract: The online detection of WO₃ Leaching Rate in tungsten hydrometallurgy is difficult to accomplish. A soft measurement plan with a convenient and fast detecting soft measurement model is devised by using MATLAB and the data collected from industrial field train neural network according to the chemical reaction mechanism of tungsten alkali hot leaching and the factors which influenced the leaching rate. Simulation results manifested the soft measurement model could well reflect actual reaction situation. The measuring relative error is less than 0.5 %. The model fit the industrial requirement. This research provides a new method for onl ine detection of tungsten leaching rate.