The concise Oxford dictionary of current English defines a model as ‘a simplified […] description of a system […] to assist calculations and predictions’. One can apply this definition, in its wider sense, to any intellectual activity (or its product) that tries to make out the components of a system and to predict the outcome of their interaction. Since models are simplified descriptions of reality, no single model will capture all aspects of a system, down to its finest components and interactions, as a ‘complete’ model would cease to be a model and turn into a copy of the original (or, in biology, a clone). Models, like tools, are developed for a particular purpose and applicable to a restricted set of tasks. This applies equally to ‘wet’ experimental and ‘dry’ computational models. A major challenge for bio-medical research is to bridge the enormous span from whole body structure and dynamics to the molecular level. This involves about 10¹⁰ orders of spatial magnitude (from metres at the organisms’ level to Angstrom for ion channel pores) and 10¹⁸ orders of temporal magnitude (from decades to describe organisms’ life spans to nano-seconds for atomic level molecular dynamics). The range of reasonable spatio-temporal combinations by far exceeds experimental and theoretical ‘modelling power’. Thus, a multitude of modular and complementary models are required. Advanced bio-engineering leads the way towards building wet experimental models that are increasingly relevant (for the questions of modern research), representative (of the system modelled) and reproducible (to allow solid characterisation of components and their interaction). Computational modelling, on the other hand, provides the way to re-integration of the pieces of the jig-saw identified using advanced experimental tools. This process involves the continuous iteration between theoretical and practical work, for model construction, validation, interpretation, and hypothesis formation. This is a bi-directional process which, if applied successfully, reduces time and resources required for biomedical research, while improving quality and applicability of findings.