The quantitative assessment of bioaerosols is important in a number of industrial and health-care applications. Measurements to aid the control of airborne infection in hospitals, detection of the release of potentially harmful microorganisms from bio-processing equipment and monitoring the exposure of workers and animals to bioaerosols are examples of situations where different concentrations of microorganisms and varying ambient conditions may be expected. Microorganisms are notoriously difficult to assess accurately under such variable conditions and no single assay method is suitable for ail applications; rather the method needs to be tailored to the application of interest. Problems are compounded by the differences in assay method (such as the type of media used for culturable counts) or sampler type selected, making the interpretation of the results difficult. An understanding of the airborne behaviour of microorganisms over a range of environmental conditions is vital if procedures are to be defined and recommended for the handling, sampling and assessment of bioaerosols. Microorganisms that are robust over a wide range of conditions are ideal as tracer particles. Unfortunately, the large majority of non-fungal bioaerosols are susceptible to damage. A predictable sampling procedure is required which will not affect the viability of the collected sample. Studies, reported on the development of procedures to characterise airborne biological particles, included tests carried out under controlled environmental conditions to compare the performance of five industrially important bioaerosol samplers with that of a reference glass wet-walled, cyclone sampler, using test aerosols of Saccharomyces cerevisiae cells and Penicillium expansum spores. The samplers tested were Andersen Microbial Sampler, Biotest RCS, Biotest RCS Plus, MicroBio MB1, and MicroBio MB2. The cyclone sampler, Andersen Microbial Sampler and the MicroBio MB1 and MB2 meet the basic criteria for a suitable reference sampler, except that the last three systems do not provide total counts. It will be important to investigate in the future how well they perform under the challenge of "real-life" conditions.