The particle size distribution of powders may have significant effects on the quality characteristics of the final product. In the case of pharmaceuticals, properties like processability, bioavailability, stability, content uniformity are influenced to a considerable extent by the particle size of the drug substances. In such instances the particle size distribution of the powders should be controlled using appropriate analytical methods andproper specifications. When acceptance criteria are laid down, the analytical and manufacturing variability should also be taken into account. This may be a challenging task since manufacturing processes, which determine the particle size characteristics, as well as the reproducibility of particle size analysis, can cause an at least an order of magnitude higher variability than usually obtained at other quality characteristics, as e.g. chemical purity. Typical manufacturing and analytical variability and their effect on the results of laser light diffraction analysis are presented. The examples deal with milling, homogeneity, quality of un-milled material, reference materials, different types or suppliers of the instruments, test material specific repeatability and reproducibility. In each case the relative deviations of the characteristic values of the particle size distributions were in the range of 20% or even more. It is obvious that the quality of a product cannot be guaranteed within a smaller acceptance range than given by the combination of all these variability causing impacts. The examples presented also show that it would be very difficult to obtain for each material a complete data set including all variability aspects. This paper discusses a capable way to set a specification without the exact knowledge about the over all variability.