The Department of Energy (DOE) is required by law to ensure that its employees work in an environment safe from hazards such as beryllium (Be), a metal that, if inhaled, can cause scarring and inflammation in the lungs. This requirement involves regularly collecting data to demonstrate that the amount of hazards present in the work environment is less than a specified OSHA limit, e.g., 0.1 micrograms per square cm for Be. Data are collected at a suspected site using premoistened “wipes,” which are sent to an independent laboratory where content is measured for Be and other toxins. In a clean environment (i.e., only naturally occurring Be), Be is typically below both the OSHA limit and the measurement system's detection limit. If Be is below the detection limit, the limit is returned, producing a left-censored measurement. Additionally, different machines may make measurements, each with their own limit. In this talk we illustrate some of the statistical approaches for Be data: analyses of univariate data sets, collecting data to understand the natural Be background, and estimating relationships between naturally occurring Be and other metals in the environment.