Although exposure monitoring has been the most common method of evaluating worker exposures, it is not necessarily the optimal method, or even a feasible method, in many instances. Exposure estimates will be most accurate when high-quality measured data are available in sufﬁcient quantities over relevant time periods. However, such data may not always be available. For instance, in many epidemiological studies, there is a need for exposure estimates of workers in processes that no longer exist. Sometimes, one may want to estimate exposures for a new process that is still in its planning stages to evaluate the adequacy of the proposed engineering controls or respirator selection. Another reason might be to prioritize monitoring efforts so that resources, which are limited, are spent on monitoring the workers at maximum risk. Thus, model outputs can act as a guide for monitoring efforts. In all these scenarios, there is a need to estimate exposures without actually carrying out monitoring. Exposure modeling is a tool that provides such estimates. In more general situations, there may be some available monitoring data, but they may be incomplete or of poor quality. In such cases, a judicious combination of monitoring data and exposure model outputs may provide better exposure estimates than any one source of information by itself.