Adverse outcomes from tobacco use may take decades to develop. Biomarkers are measures that can be used in the early stages of tobacco use to assess exposure to tobacco toxins or to predict adverse health outcomes with which they are associated. Examples of biomarkers include specific chemical components of tobacco or their metabolites; early biochemical, histological, or physiological effects; and early health effects. Mechanistically relevant and quantitatively valid biomarkers are essential for assessing the ultimate impact of new products, treatments, preventive measures, and public health policies on tobacco-related disease. The tobacco industry's recent introduction of a variety of new tobacco products or devices with implied claims of reduced health risks highlights the need to develop methods for assessing their potential for benefit or harm. A wide variety of biomarkers for tobacco exposure or harm has been studied. Although many questions about their use remain unanswered, substantial data exist regarding their validity and utility. This conference reviewed both the general issues surrounding biomarker use and the current state of knowledge regarding the most widely studied and promising biomarkers.