Human biomonitoring investigations have provided data on a wide array of chemicals in blood and urine and in other tissues and fluids such as hair and human milk. These data have prompted questions such as a) What is the relationship between levels of environmental chemicals in humans and external exposures? b) What is the baseline or "background" level against which individual levels should be compared? and c) How can internal levels be used to draw conclusions about individual and/or population health? An interdisciplinary panel was convened for a 1-day workshop in November 2004 with the charge of focusing on three specific aspects of biomonitoring: characteristics of scientifically robust biomonitoring studies, interpretation of human biomonitoring data for potential risks to human health, and communication of results, uncertainties, and limitations of biomonitoring studies. In this report we describe the recommendations of the panel.