under which exposure to it could be judged safe. EPA asked the committee to consider: (1) the conditions for which EPA should accept, consider, or rely on third-party, human toxicity studies (see Chapters 3-7); (2) under what circumstance(s), if any, the availability of human data should lead EPA to consider reducing or removing the customary 10-fold interspecies uncertainty factor (see Chapter 7); (3) the applicability of existing standards (e.g., the Common Rule, the Declaration of Helsinki) for evaluating the design and the conduct of this type of research (see Chapters 2 and 5); (4) whether and if so how the requirements of the Common Rule should be extended to the conduct of third-party human studies intended for submission to EPA in support of a regulatory decision (see Chapters 4-6); and (5) the extent to which, and how, the submitter of research with human subjects should be required to document or otherwise demonstrate compliance with appropriate standards for the protection of human research participants (see Chapters 3, 5, and 6).¹ The organization of this report has been a challenge because the issues and analysis are so intertwined. An effort has been made to provide a coherent narrative, but it has been necessary to make numerous cross-references among chapters.

The primary impetus for EPA’s request was a series of events involving agricultural pesticides and EPA’s implementation of the 1996 Food Quality and Protection Act (FQPA). This law modernized the safety standards applicable to pesticide residues in food, adding an extra measure of protection for children and placing strict deadlines on EPA’s congressionally mandated program to ensure that all agricultural pesticides currently on the market satisfy the updated safety standards. The enactment and anticipated implementation of FQPA brought into question whether current uses of certain categories of long-used pesticides—the organophosphates (OPs) and carbamates—could be maintained under the new standards.

As a general rule, EPA sets safe levels of exposure to pesticide residue in food on the basis of extensive testing in animals to determine its toxic properties and to derive a Reference Dose (RfD). It then divides the highest dose at which the most sensitive indicator of human risk did not occur (the no observed adverse effect level or NOAEL) by two or more uncertainty factors to yield the relevant RfD. One uncertainty factor accounts for the possibility that the average human could be more sensitive to the chemical’s effects than the animal model from which the NOAEL was identified (the interspecies factor). A second factor accounts for the possibility of variation among humans in their sensitivity to the chemical (the