Abstract
Although a great deal of occupational exposure data is collected, it is probably insufficient to truly answer the question of legislative compliance, ill directed in terms of real workplace risks, and is of little subsequent use for epidemiological research. This paper is an attempt to summarise the more important components and requirements of a sampling strategy, and it is therefore aimed at those with this responsibility. Perhaps, all too frequently, the more esoteric nature of these issues and their research means that they are published in journals outside the normal sphere of readership, or when it is within that sphere the quantity of statistical nomenclature and content makes it too daunting to attempt to read. By simplifying and summarising, this paper is intended to help justify a change in the sampling programme and to initiate debate.