The space thermal environment is one of the important factors that affect the image quality of a space optical remote sensor (SORS), so discussing the relationship between image quality and the temperature field of a SORS is necessary for thermal control to improve image quality. In this paper, a new thermo-optical analysis method is proposed for the optostructural system of a SORS. Thermal-elastic distortions of all components of an optostructural system, resulting from temperature field changes, cause dimensional instability of the system. This instability is described in terms of changes of position (translations along and rotations around coordinate axes), equation parameters, and roughness of optical surfaces. Then, the theory of homogeneous coordinate transformations and linear fitting of revolving conicoid surfaces are applied to processing of data from finite-element analysis to obtain all the thermally induced dimensional instabilities. Finally, these instabilities were incorporated in optical models (e.g., Zemax) to calculate the thermally induced degradation of image quality (modulation transfer function, wavefront error, etc.). The results of thermo-optical tests suggest that this method is effective.