This communication examines the role of back-calculation in developing a cleanup goal from a probabilistic risk assessment. Although back-calculation is not always appropriate with a Monte Carlo analysis, if the target risk level is specified as a single value (e.g., 95% of the population must have a cancer risk below 10−5), then back-calculation can be used to solve for a cleanup goal that represents an average concentration for an exposure area consistent with the stated target risk. This rule applies in developing screening levels and in probabilistic risk assessments that examine the influence of uncertainty in the average concentration. Back-calculation is not used to develop cleanup goals when risks arising from variable concentrations are assessed, for example, when exposure areas are very small such as for some ecological receptors, or when exposure frequency is very low such as for tourist fishermen. In this case, the cleanup goal is derived from an iterative risk calculation considering various possible truncation values of the concentration distribution. A cleanup goal derived in this manner does not correspond to a required average, but rather represents the maximum concentration that should be left in the field. Finally, although single value target risk specifications are common today, there may be advantages to setting target risks for multiple percentiles of the population, complicating the effort to calculate a cleanup goal.