For a model to be successful it must address the concerns and needs of a target audience. Further, the model must demonstrate the same patterns of behaviour as the real system, and need only be complex enough to meet this objective. If the model is too complex it can become unusable to those who need it most. With this in mind, SUPASIM has been developed for the practical metallurgist requiring only laboratory flotation rate data as input. Some 35+ successful plant simulations have been completed since 1985. The approach of SUPASIM is to characterise an ore, organise information in an understandable way and add structure to the complex flotation situation and hence support the metallurgist to interpret the relationship between variables and optimise the flotation process. The paper describes the basis on which the model was developed and sets out to provide an insight into interpreting laboratory flotation kinetics and their physical meaning in a production plant and the influence floatable gangue has on float performance and plant design. A case study illustrates how the ratio of slow floating mineral to gangue can influence circuit design and improve recovery on a plant scale. (C) 2005 Elsevier Ltd. All rights reserved.