The first public health alerts related to industrial soil contamination emerged in the early 1980s (Love Canal, Woburn, etc.). The epidemiological studies conducted to address public concerns faced two major challenges: estimating exposure and the small size of the study populations. Today, the public is demanding biological measurements because they want a real measure of their exposure. It has become necessary to clarify the decision-making process that determines the usefulness, relevance, and feasibility of conducting an exposure study using biomarkers. Three examples illustrate current practices and demonstrate the heterogeneity of the situations to be analyzed and managed. However, three key questions emerge that structure the stages of a decision-making process: Is there a benefit to proposing a health study? Is it relevant to conduct an exposure study; is the use of a biomarker the best approach? Is the study feasible, and in particular, is there a biomarker capable of meeting expectations? Answers to these three questions are essential, as the current lack of knowledge precludes rushing into this type of study without prior reflection. Drawing on its experience, the French Institute for Public Health Surveillance is currently drafting guidelines for public health professionals. (R.A.)