Measurements are reported of the short-term resuspension of nominal 10 and 20 mum alumina spheres and graphite particles from a polished stainless-steel flat plate in fully developed turbulent channel how. Preliminary measurements were made of the normal and tangential forces holding the particles onto the surface. Whilst both forces had a broad spread and were on average much reduced in value compared to that for smooth contact, the average tangential force was typically of order 1/100th of the average normal adhesive force, suggesting as has been reported previously that drag forces can play a more important role in esuspending a particle than lift forces. The resuspension measurements are compared with predictions of the RRH (1988) kinetic model based exclusively on lift/normal forces and those of a rock'n roll kinetic model that involves the rocking of a particle about an asperity (the motion being dominated by the drag force). The RRH model consistently under predicted the amount of resuspension in contrast to that of the rock'n roll model which gave values much closer to the measured resuspension. As in the RRH model, the rock'n roll model admits the possibility of removal of particles by resonant energy transfer. However the results indicate, at least for the cases considered, that this contribution is generally small, in which case the resuspension rate constant reduces to the form appropriate for a balance of moments due to adhesion and aerodynamic forces about the surface asperities at the points of contact. Under these 'quasi-static' conditions a simpler and more exact model for resuspension can be constructed: however in practice for a Gaussian distribution of removal forces this gives very similar results to the original rock'n roll model. The formula for the resuspension rate constant under 'quasi-static' conditions has similarities with the empirical formulae proposed by Wen and Kasper (1989) on the kinetics of particle re-entrainment from surfaces. Journal of Aerosol Science, 20, 483-498, although there are important differences in interpretation. (C) 2000 Elsevier Science Ltd. All rights reserved.