In new hygienically controlled plants, products often have a low level of microbial contamination, so that current methods for estimating bioburden appear to be inadequate. The adoption of efficient procedures giving consistent and reproducible results could contribute to the improvement of conventional methods for evaluating microbiological quality of products with low bioburden. The effectiveness of a washing procedure and mechanical shaking for the removal of Bacillus subtilis spores from pre-inoculated cotton gauze samples was tested in combination with a membrane filtration technique. A 45-min agitation in the presence of surfactant and glass beads improved recovery up to 70.5%, with satisfactory reproducibility. In order to compare the procedure with the current standard method, uncontaminated samples were processed to extinction by applying a repetitive treatment. When exhaustive rinses were performed in order to calculate a conversion factor, permanent entrapment of a high percentage of organisms in the cotton microfibers was highlighted: this fact may play a role in an overestimation of the extrapolated removal efficiency.