This tutorial review deals with the rapidly developing area of modelling oxide materials at the nanoscale. Top-down and bottom-up modelling approaches and currently used theoretical methods are discussed with the help of a selection of case studies. We show that the critical oxide nanoparticle size required to be beyond the scale where every atom counts to where structural and chemical properties are essentially bulk-like (the scalable regime) strongly depends on the structural and chemical parameters of the material under consideration. This oxide-dependent behaviour with respect to size has fundamental implications with respect to their modelling. Strongly ionic materials such as MgO and CeO(2), for example, start to exhibit scalable-to-bulk crystallite-like characteristics for nanoparticles consisting of about 100 ions. For such systems there exists an overlap in nanoparticle size where both top-down and bottom-up theoretical techniques can be applied and the main problem is the choice of the most suitable computational method. However, for more covalent systems such TiO(2) or SiO(2) the onset of the scalable regime is still unclear and for intermediate sized nanoparticles there exists a gap where neither bottom-up nor top-down modelling are fully adequate. In such difficult cases new efforts to design adequate models are required. Further exacerbating these fundamental methodological concerns are oxide nanosystems exhibiting complex electronic and magnetic behaviour. Due to the need for a simultaneous accurate treatment of the atomistic, electronic and spin degrees of freedom for such systems, the top-down vs. bottom-up separation is still large, and only few studies currently exist.