In this contribution, we report on the development of a similarity variable on which universal predictive models for solid hydrocarbon heat capacities at constant pressure can be based. Density Functional Theory (DFT) computations highlight the importance of elemental composition in determining heat capacity values for solid hydrocarbons and illustrate the insensitivity of specific heat capacity to the details of molecular structure over a broad range of temperatures. The computational results led to the creation of a similarity variable, based on the number of atoms per mass of a molecule. This similarity variable provides a foundation for accurate predictive heat capacity correlations for a wide range of pure and multicomponent organic solids with known or unknown molecular structures over a broad range of conditions. One such correlation and its application are described in detail in two companion contributions. (C) 2008 Elsevier B.V. All rights reserved.