It has been recognized for several decades in the pharmaceutical industry that the safety evaluation of pharmacological agents must include pharmacokinetic (PK) studies, which are designed to determine the rate of absorption, distribution, metabolism, and excretion (ADME). In recent years the importance of such ADME studies in toxicology has also become increasingly apparent to the chemical industry. This increased focus has led to testing strategies that can produce ADME/PK data with greater applicability to toxicity testing and risk assessment. An example of such a strategy is the concept of a tiered approach to the conduct of ADME/PK studies (Wilson, A. G. E., Frantz, S. W., and Keifer, L. C. (1994). Environ. Health Perspect., in press). However, in practice, PK data are often viewed as being of limited usefulness and of only ancillary importance to the determination of chemical toxicity. As a consequence, the close integration of PK studies with toxicity-testing protocols is not always practiced within the chemical industry and is thus frequently scheduled independently from toxicity testing. This lack of integration has resulted in the design of subchronic (13-week) and chronic (2-year) toxicity studies without the benefit of PK information to establish the appropriate dose levels to be used, often because of inappropriate timing. The result is that much of the PK data which have been generated is without a clear consideration of its application to toxicity testing and risk assessment. This position paper is intended to provide recommendations for the appropriate design and interpretation of a PK study, as well as when and how to use PK data in the interpretation of toxicology data. Additional issues discussed in the paper include the design of PK studies to evaluate tissue time-course relationships and chemical persistence, the overall usefulness of PK data to toxicology testing, and the utility of PK as a useful interpretive and predictive tool in toxicology and risk assessment.