Stirred milling technology has become an indispensable process in the mineral processing industry. Many ore deposits classified in the past as "untreatable" have been economically extracted enabled by improvements in this technology. Gravity induced stirred mills are widely used in the industry, but the effect of classification above the grinding zone (referred as internal classification zone) is not well understood. The authors have conducted research to develop a process model for the gravity induced stirred mill. An integral part of this study is to understand and develop a predictive model for the internal classification zone above the grinding region. A rising bed classifier has been designed, and a series of experiments were carried out by varying the slurry solids concentration and flow velocity to replicate the gravity induced stirred mill classification effect. The results show that the internal classification effect is significant when the slurry is dilute and flow velocity in the mill is low. The cut size (d(50C)) coarsened from 158 mu m to 2040 mu m and sharpness of the cut (alpha) increased from 0.18 to 0.84 with a rise in slurry solids concentration (v/v) from 26% to 46% and flow velocity from 10 mints to 21 mm/s. The model developed from this data can be combined with the population balance model to simulate the internal classifier effect of an industrial scale gravity induced stirred mill. (C) 2016 Elsevier Ltd. All rights reserved.