The calculation of exposures to contaminated soil at Superfund sites has traditionally involved a steady state approach (the time-weighted average intake) which is uncalibrated and unvalidated, based on exposure factors which may be derived from risk management decisions, and which fails to take pharmacokinetics into account. Although the approach has some attractiveness from the standpoint of amenability to standardization and simplicity of operation, it has been criticized as being overly conservative, i.e. it yields chronic intake values which far exceed those actually measured through biological monitoring. Recently, USEPA has developed and published an alternative approach, known as the Integrated Uptake/Biokinetic (UBK) model for the determination of exposure to environmental lead. This model has two principal components -- one involving the uptake of lead into the body from different sources, the second involving the calculation of a blood lead level from the uptake. The blood lead level is subsequently used in conjunction with dose-response functions to predict the probability of adverse health effects. Unlike the average intake approach, it has undergone calibration and validation, is relatively free from risk management, and incorporates pharmacokinetics. The purpose of this paper is to present the results of a critical evaluation of the different approaches used to calculate exposure to lead.