Top-spray fluid bed coating scale-up experiments have been performed in three scales in order to test the validity of two parameters as possible scaling parameters: The drying force and the relative droplet size. The aim was to be able to reproduce the degree of agglomeration as well as the mechanical properties of the coated granules across scale. Two types of placebo enzyme granule cores were tested being non-porous glass ballotini cores (180-350 mu m) and low porosity sodium sulphate cores (180-350 mu m). Both types of core materials were coated with aqueous solutions of Na2SO4 using Dextrin as binder. Coating experiments were repeated for various drying force and relative droplet size values in three top-spray fluid bed scales being a small-scale (Type: GEA Aeromatic-Fielder Strea-1), medium-scale (Type: Niro MP-1) and large-scale (Type: GEA MP-2/3). The tendency of agglomeration was assessed in terms of particle size fractions larger than 425 mu m determined by sieve analysis. Results indicated that the particle size distribution may be reproduced across scale with statistical valid precision by keeping the drying force and the relative droplet size constant across scale. It is also shown that none of the two parameters alone may be used for successful sealing. Morphology and microscope studies indicated that the coating layer is homogenous and has similar structures across scale only when both the drying force and the relative droplet size were fixed. Impact and attrition tests indicated that it is possible to produce granules with similar attrition and impact strength across scale and that the two types of mechanical properties are inversely related. (c) 2007 Elsevier B.V. All rights reserved.