The Floatex density separator (FDS) is an industrial separator and works on the principle of hindered settling where the settling rate of a particle in suspension is affected by nearby particles. This phenomenon of hindered settling has been described by various authors using the concept of particle slip velocity. The feasibility of one such slip velocity model proposed by Galvin et al. to predict the separation of chromite in a plant scale FDS was examined. Previously this model was validated only at lab scale with synthetic mixture of various density particles. In order to use this model, the feed chromite ore was characterized and quantified into different density mineral classes and their percentages were estimated by using mineralogical grain count method. The performance of FDS was then predicted using slip velocity model in terms of weight recoveries and composition of different minerals in the FDS underflow product. The data predicted by chosen slip velocity model were validated using the plant data. The results predicted from the model were with in 5% error match with experimental results at low teeter water velocities. Therefore the slip velocity model was further explored to study the effect of teeter water velocity on the chromite grade and overall chromite metal recoveries in the applicable range of teeter water velocities. This has resulted in a fair understanding of the separation behaviour of the chromite mineral in FDS and helped in choosing appropriate teeter water velocities for production of desired grades of chromite concentrate as FDS underflow. (c) 2006 Elsevier B.V. All rights reserved.