A methodology to determine risk as a function of time varying concentration was developed and applied to three leaking underground storage tank sites for five remediation scenarios, to demonstrate the risk reduction associated with corrective actions and the transient nature of risk. The conventional method for calculating risk-based clean-up levels employs a steady-state approximation such that the receptor is assumed to be exposed to a constant concentration for the entire exposure duration. Analysis by the transient method suggests that regulated clean-up levels are conservative, and that clean-up levels derived using the conventional risk-based approach were arbitrary and highly subjective with respect to the definition of 'acceptable' risk. When compared to results from the transient analysis, the clean-up levels determined from the conventional steady-state approach were shown to be conservative by as much as an order of magnitude.