Predictive modelling of solid-liquid separation can greatly assist the design and operation of thickening and filtration equipment, improving water recovery and reducing costs. A phenomenological model describing continuous thickening has been previously developed with primary inputs being the material properties, (compressive yield stress and hindered settling function) derived from routine laboratory batch settling and filtration tests. This work aimed to validate the model by operating a pilot column continuously and measuring the underflow solids. The column was operated at two different solid fluxes and several bed heights. Additionally, the influence of flocculation conditions (polymer dosage and residence time) on thickening performance were studied. The model predicted the experimental underflow solids concentration at a given flux. For the observed underflow solids concentration, the ratio of the actual to predicted flux was observed to be between a factor of 1 (accurate) and 10. The model was most accurate for the lowest bed heights. This work confirmed the model was able to correctly predict trends for the case where minimal bed height and shear forces are present. Deviation from the model is postulated to be due to changes in toe dewatering properties of flocculated aggregates over time that are not adequately captured using conventional batch sedimentation tests. The data from these tests are traditionally used as a key input to thickening models. (C) 2010 Elsevier Ltd. All rights reserved.