Ion mobility spectrometry is a separation technique that also provides structural information about the analytes in the form of collision cross-sections (CCSs), which can be utilised for modelling. The dominant paradigm is to calculate theoretical CCSs for a range of candidate structure models and compare with experimentally derived CCS, thereby filtering out candidates not matching the experiment; however, new knowledge-based approaches are emerging. Herein, we give an overview of the concepts and the important considerations for CCS-based modelling, for analyte classes ranging from small molecules to large macromolecular complexes, and how the CCSs fit with other experimental data. Both the calculations and the experiments have specific uncertainties, and the experimental conditions might cause perturbations to the structure, complicating the interpretation. Generation of appropriate candidate structures stands out as a critical component determining the maximum accuracy of the modelling. Molecular dynamics simulations and other modelling tools can capitalise on system-specific information to explore the right parts of conformational space, while advancing our general understanding of how molecules are affected by the experiments can refine the candidate structures as well as connect CCS-based structures more closely with in vitro and in vivo chemistry and biology.