Extrinsic staining of teeth is a common aesthetic problem for which toothpaste manufacturers strive to produce effective products to chemically and/or physically control the problem. Clinical studies proving efficacy are rare and reliance on tests in vitro are more common. The aim of these studies was to develop two in vitro models which could be used to screen and compare agents and formulations for dietary stain control. The first model utilised immobilised hydroxyapatite (RA) with tea as the chromogen. Using residual tea optical density to assess extrinsic staining proved unreliable. Tea extraction and direct measurement from wells however was shown to be reliable in demonstrating the efficacy of known active chemical agents and differentiating between toothpastes. The second model used roughened Perspex to simultaneously assess toothpastes and their ingredients for chemical and or mechanical stain removal. In line with the RA model, results differentiated between toothpastes and other agents for chemical stain removal. Chemical/mechanical stain removal in the model also showed differences for toothpastes, abrasives and detergents but abrasivity levels did not show the expected linear correlation: the toothbrush itself appeared to be the dominant factor in the model. The final experiment attempted to determine whether chromogen/pellicle interaction was specific or non-specific. Results showed no evidence of specificity indeed data suggested that chromogens may merely be absorbed into the pellicle and held physically rather than by chemical interactions. Since some of the data from both models was consistent with other published laboratory and clinical studies it is cautiously concluded that the models may prove rapid, simple and useful screening methods for anti-staining agents and products. Only the availability of more clinical data to support claims for numerous products, however, will allow the models to be retrospectively validated.