The deflection that a lifter bar is subjected to when passing through a grinding charge was measured using an embedded strain gauge sensor. The obtained signal profile is analysed and interpreted in relation to charge properties such as toe and charge angle for the grinding charge. The current work has focused on the charge and slurry flow behaviour when solids content is changed and how well the sensor reflects this. Bench scale measurements with a vane type viscometer, roughly evaluated in terms of apparent viscosity is used as a character for the resistance to motion of the slurry. The slurry flow resistance is strongly influenced by solids content and obviously by the addition of a dispersant. The strain gauge sensor reflected this change well, showing that toe and shoulder region of the charge varied in a systematic way. Results obtained also shows that change of slurry flow resistance exert an influence on grinding performance. A multivariate statistical method, partial least squares regression, is applied to the sensor data producing a model that can predict the change in slurry flow resistance. The output from the model also shows good properties to be used as a process-monitoring tool. The predictive capability of the model is believed to be of such quality that it can be used for process control. (c) 2005 Elsevier Ltd. All rights reserved.