Continuous powder mixers offer a viable alternative to batch processes, but have received very little attention in scientific literature and in the industrial world. Mixer design is still very empirical and is not based on assessed methodologies. In this paper, we report experiments that aimed to compare two very different types of stirrers for a pilot-scale continuous powder mixer, and for two types of mixtures: a model mixture and a real pharmaceutical mixture. The first stirrer A is of the frame type with inclined paddles and internal transporting screw, the other stirrer B is of the shaft type with paddles mounted on it. Results are first presented from the viewpoint of bulk powder flow by hold-up determination and correlation with operating conditions. General relationships are derived which show that the mobile B leads to higher hold-ups, which may be an important drawback. The study of mixture homogeneity globally confirms these findings, especially in a dense phase flow regime. in the fluidised regime, where the stirrer B can be used, attention is drawn to the negative effect of excessive rotational speeds on the quality of the mixtures. (c) 2008 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.