Reference doses (RƒD) for the definition of tolerable food residues have traditionally been based on the lowest no observed adverse effect level (NOAEL), which usually comes from chronic toxicity studies. While this is generally agreed to be a safe approach to evaluate the overall significance of expected consumption it is clear that it is not appropriate for evaluation of the toxicological significance of residues in a single meal. Standard acute toxicity tests are not designed to generate an NOAEL, from which an RƒD can be derived. They are more appropriate to evaluating risk following accidental high exposure to the product itself rather than to food residues. A typical toxicological database for a pesticide active substance contains studies which may be appropriate, on a case-by-case basis, to evaluate shorter term endpoints of interest for specific molecules, such as developmental or acute neurotoxicity studies. However, their specificity limits their scope of application. General toxicological endpoints are well covered by short-term, 28- or 90-day, guideline studies. However, neither of these studies is ideal for setting of an acute RƒD (ARƒD) as the treatment period is significantly longer than the duration of consumer exposure. This could be balanced by applying a reduced safety factor to the NOAEL to set the ARƒD. Alternatively, a test guideline could be designed to generate a relevant acute NOAEL but the time necessary for development, validation and acceptance of such a guideline means that an interim approach is, in any case, necessary.