The significance of excursions of intake above the ADI, TDI, or PTWI can only be assessed by reference to the database which led to the derivation of these, most particularly the duration of the pivotal study (chronic, subchronic, acute), the pharmacokinetic parameters, and the nature of toxicity and mechanism of action. Although this implies a case by case assessment, a number of typical situations may be recognized: (1) The substance (usually a contaminant, not an additive) has a very long half-life leading to accumulation in target organs/tissues, e.g., Cd or dioxin. The chronic toxicity is manifested when critical concentrations are achieved in these tissues and there is a large difference between the acutely toxic dose and the chronic NOAEL. In such a case, the effect of excursions above the PTWI on tissue levels is readily calculated; peak excursions of several times the PTWI for short periods (days, weeks, or even months) or lower peak intakes for even longer periods (months to years) may be inconsequential provided that the integrated exposure over longer periods does not lead to critical steady-state tissue concentrations being achieved. (In such cases, it is clearly inappropriate to divide the PTWI by 7 and treat this as an ADI.) (2) A more common situation for food additives is where the ADI is based on a chronic study, but the t1/2 is short, i.e., the situation is one of chronic stress rather than cumulative toxicity. In such cases, e.g., BHA or BHT, the effects on the target organ (hyperplasia, foci of altered cells, etc.) can usually be identified in subchronic studies, although progressive changes may occur in chronic studies. Two subsituations then arise. First, when the effects seen at the LOAEL in subacute/subchronic studies are truly reversible (e.g., methemoglobinemia), short-term studies with a reversibility component may become pivotal in assessing the consequences of short-term excursions above the ADI. Second, when the short-term effects are not fully reversible, or are even progressive, the consequences of short-term peaks of intake above the ADI would require careful evaluation against the NOAEL or LOAEL in subacute or subchronic studies. (3) A rare situation might arise where the ADI is based on a chronic toxicity study but the margins between the chronic NOAEL and some aspects of acute toxicity may be small. For example, for a compound which behaved like retinol, the ADI might be based on chronic effects on the liver but at maternally nontoxic doses the substance may be teratogenic following acute exposure during early pregnancy. Clearly in such a situation, the acute NOAEL for teratogenicity would be used appropriately to evaluate the risks associated with short-term peaks of exposure, i.e., a different study may be pivotal in determining the effects of large excursions above the ADI than that which was used to calculate it. Clearly, these cases are not comprehensive but do provide a framework against which to discuss the potential effects of excursions above the ADI and to reach rational conclusions which are not based on the misapprehension that the ADI (or worse, the PTWI x 7) is a lower bound of toxicity.