An analysis was conducted on the role played by biological markers (biomarkers) in toxicant exposure assessment, and also on the criteria for using biomarkers in environmental epidemiological studies. Biomarkers can be used to determine four levels of association between exposure and response: presence of response, shape of the exposure/response curve, threshold or no effect level, and attributable risk over the entire exposure range. Valid biomarkers of exposure were dose markers having biological, pharmacokinetic and temporal relevance, defined background variability, and identification of confounders. The development of individual biomarkers followed a progressive series of experimentally derived measurements that determined whether the biomarker responds appropriately to various exposure situations. Following laboratory validation, the biomarker was subjected to field studies to determine the extent to which it varies by demographic, geographic and other relevant descriptors. A biomarker decision model has been developed to determine population exposure levels. The ultimate test of biomarker utility was whether or not it can be used to reliably predict disease development in humans. Exposure intervention studies can be effective in the evaluation of a biomarker and the intervention upon validation of the biomarker.