The development of medicines during the 20th Century was initially based on oral delivery of drugs via the gastrointestinal tract. To enhance understanding of rate of uptake of different drugs and formulations and reduce the need for animal testing, in vitro models based on gut epithelial cell models were developed in the 1980s and 1990s. With the advent of biotechnology, an increasing number of drugs based on proteins and other biomolecules are being produced, which currently require parenteral administration (by injection). To avoid the need for injection, alternative routes of delivery are being sought for these molecules, including mucosal routes of the gastrointestinal tract and the lung. In parallel with this, the field of ‘nanotechnology’ began to develop. Nanotechnology offers both solutions and problems. ‘Nanomedicines’ over a range of nano sizes appear to offer some solutions for delivery, provided that they could cross epithelial barriers. In contrast, there remains considerable concern that the many different types of nanoparticles in development for electronics and new materials may be taken up into the body and cause harm. There are therefore clear needs for epithelial models which allow us to not only screen conventional drugs for absorption, but also assess potential non-invasive delivery of biologics and nanomedicines, as well as screen easily and reliably for nanotoxicology1. As it is the same barrier involved in all of these cases, we need a single epithelial model that can adequately reflect and give accurate answers for all of these different barrier problems. In this article, we assess the properties needed for an epithelial cellular model, the current state of the art, and some recent work developing a more accurate and comprehensive model.