Although it has been recognised for many years that biological particulate matter in the atmospheric environment can trigger symptoms of allergic respiratory diseases such as asthma, the results of studies examining the relationships between pollen counts and the occurrence of such diseases have been inconsistent. In addition to the size of pollen grains as an explanation for such disagreement between studies, their is now a body of literature which has demonstrated that airborne pollen allergen can exist in sub-pollen sizes and out of the 'pollen season', and that little correlation can occur between allergen levels and pollen counts. These findings not only explain disagreement between epidemiological studies using pollen counts but may raise doubts over the plausibility of any results from such studies. The paper reviews the results of a selection of epidemiological studies of pollen counts and asthma as well as studies which have documented the existence of pollen-free airborne allergen. It is concluded that future epidemiological studies should measure allergen rather than pollen grain counts, using methods that have been developed specifically for this purpose. Further research is required to determine if the presence of airborne pollen-free allergen is a universal phenomenon in pollens and perhaps in moulds as well.