This study presents an evaluation of the predicted and experimentally observed flow behavior, i.e., flow pattern and critical outlet diameter, in a tablet press hopper. Three grades of a commonly used pharmaceutical additive, microcrystalline cellulose, were used for this purpose. The prediction of the flow behavior was made by first extrapolating shear and wall friction data and, subsequently, applying these extrapolation-based data as input variables to Jenike's hopper design method. Predictions were then compared with empirical discharge experiments from a tablet press hopper and a funnel with varying outlet diameters. By application of a mixed effect 2nd order polynomial model for the extrapolation of the wall yield locus, accurate flow pattern predictions for the tablet press hopper was made. However, the observed flow patterns were very sensitive to the powders' moisture content. The comparison of the observed and predicted critical outlet diameters showed good agreement for the powder with the best flowability when linear extrapolation of the flow function was applied. In contrast, the predicted critical outlet diameter was slightly overestimated compared to the experimentally observed diameter for the two more cohesive powders. A likely reason for this overestimation is that the flow function probably has a non-linear convex upward shape for these two powders at very small consolidation stresses. These findings illustrate the relevance of measuring shear and wall shear stresses at very small consolidation stresses to improve the flow behavior predictions for small-scale process equipment typically used during production of solid state pharmaceuticals. (C) 2017 Elsevier B.V. All rights reserved.