Biomonitoring, the measurement of chemicals in human tissues and fluids, is becoming commonplace, and biomonitoring data has proved to be an important resource for identifying the presence of chemicals, both natural and synthetic, in human populations. However, the concentrations of the chemicals detected in human samples are generally very low, typically in the parts per billion (ppb) or parts per trillion (ppt) range, and the degree of risk posed by these chemicals depends on whether the exposure levels approach those known to cause toxicity in test animals or people. Unfortunately, it is often difficult to relate a measured concentration of a chemical in a human tissue or fluid to the administered doses used in animal toxicity studies. As the number of chemicals identified in human tissues increases, so does the challenge for providing a risk context for the observed concentrations. Moreover, the challenges associated with interpretation of biomonitoring data on different classes of chemicals can be quite different. This review focuses on the use of pharmacokinetic modeling, and in particular, physiologically based pharmacokinetic (PBPK) modeling, to support the interpretation of human biomonitoring data from the perspective of exposure reconstruction and risk characterization. A general approach, referred to as reverse dosimetry, is described for estimating the distribution of exposure levels in the environment that could give rise to measured biomarker concentrations in a population. These exposure distributions can be compared to regulatory exposure guidance values or no-effect levels in toxicity studies to put potential risks in context.