With a view to establishing a model system for examining toxicity in skin, primary lingual keratinocyte cultures , a keratinocyte cell line and freshly isolated keratinocytes all derived from rat sub-lingual epithelium, were partially characterized both morphologically and enzymically and the toxicity and metabolism of the mycotoxin T-2, studied therein. A number of techniques for obtaining pure suspensions of sublingual keratinocytes (NCK) were examined and dispase is recommended for completely separating the epithelium from its dermis prior to trypsinization. Existing methods for the culture of PLK cells were improved to reduce the number of rats used and minimize the fibroblast contamination and a technique for culturing the epithelium associated with human hair follicles was also examined. The follicle technique was found to produce primary cultures which were 100% epithelial but considerable time and resources were required to generate relatively few cells. The keratinocyte cultures, PLK and RTE5, were shown to produce keratin and undergo stratification like the epithelium in vivo, Using a series of specific enzyme inhibitors, both the cultured and non-cultured epithelial cells were found to possess the same characteristic forms of acetylcholinesterase, butyrylcholinesterase and carboxylesterase. The effect of T-2 on protein synthesis in the keratinocyte cells was examined and a dose-related inhibition was evident. The primary and freshly isolated cells appeared to be the most resistant to synthesis inhibition. The rate of recovery from this inhibition was greatest in the cell line which also lost T-2 at the fastest rate. The uptake into and loss of T-2 from the keratinocyte cells was chiefly by simple diffusion, but studies using rotenone demonstrated that an active process may have been involved in reducing the rates of uptake and loss. The total amount of T-2 absorbed was likely to be dependent on the number of binding sites while the overall rate of loss was dependent on the strength of that binding. The possible nature of these binding sites is discussed. T-2 was metabolized in all the keratinocyte preparations to the same products found in mammalian skin, in vivo, and studies with specific enzyme inhibitors showed that a carboxylesterase was most likely to be involved in its hydrolysis. The greater metabolism of T-2 in the primary cells, which contained the most carboxylesterase, is likely to have been one of the factors involved in reducing the levels of protein synthesis inhibition in these cells. Rat sublingual epithelium and the cultures derived from it have a similar morphology and esterolytic capability to that of skin in viva. The PLK cells were most similar to NCK cells in terms of protein synthesis and esterolytic capability, so if it is assumed that NCK cells are representative of the sublingual epithelial cells in viva and hence skin, then the primary cultures might prove to be the best model culture system for studying toxicity in skin. Nevertheless, the RTE5 cell line was most similar to the NCK cells in respect of T-2 metabolism, and if it shown to have other characteristic skin metabolism systems then it too might prove useful for studying skin toxicity. In addition, the cell line would probably prove to be a more convenient, simpler, reproducible and cheaper means of examining skin toxicity on a large scale.