During the operation of high-speed trains, components are prone to fatigue crack damage under the action of cyclic loading and extreme temperature environments. Temperature change has become an important factor for monitoring the health of high-speed train structures. A new fatigue crack size quantification method under a variable temperature environment based on a Gaussian mixture model (GMM) is presented in this paper. A series of damage indexes are proposed to characterize the interaction mechanism between the signal and crack under temperature change. Moreover, multidimensional damage indexes extracted by the fusion of dimensionality reduction technology are used to establish a complete working condition baseline GMM database. The service temperature of the structure is determined according to the maximum similarity criterion between the baseline GMMs and detection GMMs. Furthermore, the quantitative crack length detection model is established at each temperature. To validate the effectiveness of the method, fatigue crack experiments in a variable temperature environment are carried out. The verification results show that this method can detect fatigue crack growth at different temperatures.