Methionine (Met) is the primary limiting essential amino acid in grain legumes. The imbalance in amino acid composition restricts their biological value (BV) to 55 to 75% of that of animal protein. So far improvement of the BV could not be achieved by conventional breeding. Therefore, genetic engineering was employed by several laboratories to resolve the problem. Three strategies have been followed. A) Engineering for increased free Met levels; B) engineering of endogenous storage proteins with increased numbers of Met residues; C) transfer of foreign genes encoding Met-rich proteins, e.g. the Brazil nut 2S albumin (BNA) and its homologue from sunflower, into grain legumes. The latter strategy turned out to be most promising. In all cases the gene was put under the control of a developmentally regulated seed specific promoter and transferred into grain legumes using the bacterial Agrobacterium tumefaciens-system. Integration into and copy numbers in the plant genome as well as Mendelian inheritance and gene dosage effects were verified. After correct precursor processing the mature 2S albumin was intracellularly deposited in protein bodies which are part of the vacuolar compartment. The foreign protein amounted to 5 to 10% of the total seed protein in the best transgenic lines of narbon bean (Vicia narbonensis L., used in the authors' laboratories), lupins (Lupinus angustifolius L., used in CSIRO, Australia), and soybean (Glycine max (L.) Merr., used by Pioneer Hi-Bred, Inc., USA). In the narbon bean the increase of Met was directly related to the amount of 2S albumin in the transgenic seeds, but in soybean it remained below the theoretically expected value. Nevertheless, trangenic soybean reached 100%, whereas narbon bean and lupins reached approximately 80% of the FAO-standard for nutritionally balanced food proteins. These results document that the Met problem of grain legumes can be resolved by genetic engineering.