Actual diagnostic criteria in food allergy frequently show a group of subjects with uncertain diagnosis, i.e., subjects with only occasional reactions to an offending food. These subjects are probably under- or misdiagnosed. It is possible that food itself may be only one of the factors involved in the allergic food reaction, at least in these subjects. In these cases, several factors besides food could modify either the allergenic potential of food or the absorption of allergens and may be relevant for a clinical response to occur. In order to investigate the possible relevance of missing factors, a theoretical model is elaborated by modifying the equation that represents the classical food allergy model: xAg+( IgE ∗ -cell)⇒ clinical response (mediators release…). In this classical model, it is assumed that if there is specific IgE against a food (IgE*-cell), the only limiting factor will be x (total amount of allergen ingested), being Ag, the allergen supposed to elicit a clinically relevant response. Always that x⩾ a (threshold quantity of allergen that elicits a reaction) a reaction should occur. The possible existence of missing factors, i.e., the presence of ‘modified allergens’ ( Ag ′, Ag ′′,…, Ag n ) induced by modification of the allergenic potential of food, are not considered in the equation, but these factors may have clinical relevance, and these relevance could be related to x 1, x 2,…, x n (relative quantities of ‘modified allergens’). So, a modified formula is proposed: {x−Σ(x 1,x 2,…,x n)]Ag+Σ(x 1Ag ′,x 1Ag ′′,…,x nAg n)}+ IgE ∗ -cell)⇒ clinical response . In this formula, depending of the relative amount of the limiting factors: x 1, x 2,…, x n and the relevant allergen ( Ag, Ag ′, Ag ′′… or Ag n ) to which the subject is sensitized an allergic reaction could or could not be elicited. So, factors other that the food itself and the threshold of sensitivity may be missed in the diagnosis of food allergy. When there are not any factors modifying Ag, the allergic response will depend only on the amount of x (‘classical equation’). But when there is some kind of modification the response will depend on the relevance of the modified allergens and on the relative amount of them. If the factors modifying the allergenicity cannot be controlled, the amount of the clinically relevant allergen cannot be estimated, and the clinical response may be only occasional, explaining the inconsistent findings of some subjects.