The four IHHWP patients with associated food allergy displayed similar clinical symptoms to those of WDEIA patients, following wheat food intake. Both types of patients displayed generalized urticaria or anaphylaxis and they generally involved anti-ω-gliadin IgE. The main differences were the negativity of SPT with normal unmodified wheat products, their positivity with HWP and the apparent absence of associated effort, in the cases of IHHWP patients. This study is a further attempt to discriminate both diseases on the basis of the specificity of the IgE they involved. Recently (17) we showed that an IHHWP patient reacted like WDEIA patients with γ3-hordein, a barley prolamin cross-reacting with ω5-gliadins (16). For this reason, bread wheat varieties used to characterize the IgE were chosen to be representative of the ω5-gliadin alleles expressed in French bread wheat varieties and to which the French population has been exposed during the past 20 years. The objective was to screen the different alleles of ω-gliadins and more particularly of ω5-gliadins, for the extent of their implication in IHHWP. Among the selected varieties, Clément (no. 11), was introduced as representative of wheat varieties that present 1BL/1RS translocation and so are deprived of ω5-gliadins. These last varieties express in their place secalins (Gli-B1l alleles), which are similar to ω-gliadins. Omega gliadins of the three genomes expressed by all the selected varieties, along with their corresponding encoding loci on wheat chromosomes, could be distinguished with accuracy, according to both their relative mobility and to the characteristic band pattern they displayed upon SDS-PAGE. These patterns were the result of constitutive co-expression of gliadins inherited all together, in the grains. The reproducibility of these patterns is a basis of wheat variety identification. The combination of these gliadin properties with an accurate protein transfer and the possibility to localize accurately the chemiluminescent signal on films by superimposition on the membranes poststained for total proteins, allowed a nonambiguous identification of the IgE reactive alleles. The results showed that none of the IgE of the four patients tested sensitized to HWP reacted with ω5-gliadins. The reactive ω-gliadins were encoded mainly at Gli-D1 and Gli-B1 loci. Most of the alleles of ω-gliadins and γ-gliadins encoded at Gli-A1 did not react. Only the alleles k1, k2 and m1, m2, m3, expressed in the varieties Soisson (no. 2), Ruso (no. 10) and Clément (no. 11), were reactive with IgE. This shows the singularity of these alleles among the others which are also encoded on chromosome 1A. Concerning the Gli-B1 alleles reacting to IHHWP sera, they reacted in most cases as a single band in the γ-gliadin area. This result is in accordance with our previous observation that the serum from a patient with IHHWP, also reacted with γ3-hordein, which was homologous to wheat γ-gliadins (17). The Gli-B1 locus is known to encode both γ- and ω-gliadins. It is not known if these γ-gliadins share common epitopes with the reacting ω-gliadins encoded by Gli-D1, or if they are different antigens reacting with different IgE clones. It is interesting to remind here that in addition to barley hordeins, IgE from IHHWP patients also reacts with secalins from rye (17) (and additional results not shown). This is confirmed here by the reactions of IgE from only IHHWP patients with the Gli-B1l alleles, which are rye secalins and which replace ω5-gliadins, in the variety Clément (no. 11). This suggests that secalins structurally related to ω1,2-gliadins or to γ-gliadins encoded respectively by Gli-D1 and Gli-B1, also exist in rye. As IgE from WDEIA patients did not react to these Gli-B1l alleles (21), these homologies must be distinguished from those between ω5-gliadins, γ3-hordeins, γ35- and γ75-secalins which were already revealed by using IgE from WDEIA patients (16, 17, 21). These observations show that wheat, rye and barley could be potentially harmful not only to WDEIA patients but also to IHHWP patients. These plants are phylogenetically related, and the immunological cross-reactions are normal. Prolamins from these cereals share general common structural features, and strong immunological homologies of these proteins, within species and between species, which are different from those cited above, are known since a long time. This makes highly difficult to raise specific antibodies reacting with specific components of prolamins. Compared with the IgG raised experimentally, IgE from both IHHWP and WDEIA patients exhibited exceptional high specificities. They were able to discriminate between strongly related gliadins encoded on homoeologous chromosomes, like ω5- and ω1,2-gliadins. This shows that the epitopes against which IgE antibodies are elicited in IHHWP and WDEIA, are probably different from the common epitopes which are tolerated in normal individuals eating wheat products (22, 23). The involvement of these IgE epitopes is probably the result from several factors. Among them, modifications of the structure of the antigens, exposing buried structures, or modifications in their presentation to the immune system other than by the oral route, are possible factors. In the present case of IHHWP, both these factors occurred. The wheat protein antigens have been broken by partial hydrolysis, leading to the exposition of internal structures and to random aggregation of the fragments, as already observed (4). These modifications, combined with the application on the skin of these new antigens, more than their oral consumption, were probably determinative in the orientation of the immune system to an IgE response. Allergic sensitizations following epicutaneous exposure to proteins are now recognized (24, 25).