Cycloaliphatic epoxies (CEs) are commonly used for structural applications requiring improved resistance to elevated temperatures, UV radiation, and moisture relative to other epoxy materials. Accurate and efficient computational models can greatly facilitate the development of CE‐based composite materials for applications such as Aluminum Conductor Composite Core high‐voltage power lines. In this study, a new multiscale modeling method is developed for CE resins and composite materials to efficiently predict thermal properties (glass‐transition temperature, thermal expansion coefficient, and thermal conductivity). The predictions are compared to experimental data, and the results indicate that the multiscale modeling method can accurately predict thermal properties for CE‐based materials. For 85% crosslink densities, the predicted glass‐transition temperature, thermal expansion coefficient, and thermal conductivity are 279 °C, 109 ppm °C−1, 0.24 W m−1 K−1, respectively. Thus, this multiscale modeling method can be used for the future development of improved CE composite materials for thermal applications. © 2018 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2018, 135, 46371.