Understanding proteins well enough to rationally modify their biological function requires understanding how these biomolecules behave in their natural cellular, or extra-cellular, environments. This, in turn, implies understanding their interactions with a wide variety of other species within a dense and heterogeneous medium. Molecular modelling and simulation can certainly contribute to improving our understanding in this area, however the range of molecules and processes involved in the biological systems implies that a range of modelling techniques will have to be applied, balancing the requirements for accuracy and precision against the constraints imposed by the size, time and energy scales involved. This article attempts to summarize the various representations, methodologies and target functions presently available to molecular modellers, discusses how different combinations of these basic features can be combined to attack different problems and then considers the role of hybrid methods, an area where there is still much scope for development.