For most foods, true standardization is not yet feasible because there is insufficient information on the relative importance of individual allergens and their variants. Standardization without sufficient information may easily be counterproductive because improvements are less likely to be implemented. In the analysis of natural test material, the following principles apply: 1) RAST inhibition is usually inferior to other means of allergen quantitation. 2) Immunoblot is inefficient for some "important" allergens and overefficient for some "unimportant" allergens and may therefore be deceptive. 3) Single-component assays are the only satisfactory way to describe complex mixtures. Improving the actual food-testing procedure is important, but will not alone result in a reliable diagnostic procedure. Tests for measuring "effect modifiers" will have to be developed in order to predict in vivo reactions to foods.