In the literature on mechanically agitated solid-liquid systems, several methods are described to estimate the minimum impeller speed N-js at which all particles are suspended, but few studies have been devoted so far to their critical comparative assessment [67]. In the present paper, several alternative N-js prediction methods are applied to CFD results obtained for selected test cases covering a broad range of suspension conditions and impeller speeds. Results are compared with one another and with classic empirical correlations [88]. The aim of the work is to assess the adequacy of different methods for predicting N-js and, more generally, to contribute to a viable CFD-based strategy for the design of solid-liquid mixing equipment. To this purpose. transient RANS simulations using the sliding grid algorithm were carried out. An Unsuspended Solids Criterion (USC) was introduced to judge whether the solids contained in a generic control volume should be regarded as suspended or unsuspended. Based on this criterion, the concept of impeller speed for sufficient suspension, N-ss, was proposed. The results suggest that it may be convenient to base the design of solid-liquid contactors on the sufficient suspension speed N-ss rather than on the traditional N-js concept. (C) 2012 Elsevier B.V. All rights reserved.