The interactions between (Brady)Rhizobium and legume plants involves many interesting problems. In the last ten years, there were remarkable experiments which have detected excreted flavonoid compounds at pmol levels from plant roots, which induce (Brady)Rhizobium nod gene expression (Long 1989, Nap and Bisseling 1990, Dénariéet al. 1992, Schlaman et al. 1992). The responses of rhizobial genes to the various kinds of chemical compound are different (Maxwell et al. 1989, Zaat et al. 1989, Davis and Johnston 1990, Hartwig et al. 1990, Hungria et al. 1992). The resolution of pSym genes controlling those mechanisms makes way for the long-term goal of introducing nitrogen fixation ability into nonlegume plants. Recently, some experiments have shown that Rhizobium and other nitrogen fixing bacteria form nodule-like strutures on rice, barley or wheat (Al-Mallah 1989, Jing et al. 1990, Rolfe and Bender 1991). Some O2 protection mechanism instead of leghemoglobin must be needed for nitrogen fixation by Rhizobium or other N2-fixing bacteria which have invaded in the nonlegume root tissue. The isolation of the plant mutants or preparation of transgenic plants capable of hyper-nodule formation having efficient nitrogen fixation ability may be major goals. For the attainment of these goals, transformation of a foreign genome (nif-or nod gene cassette) into the plant cell might be a good way to proceed (Barker et al. 1990). It is also necessary to clarify the relationships between the level of relative endogenous plant hormones and the exchange of the differentiation of the root tissue to the nodule tissue. This phenomenon of redifferentiation of plant tissue by the results from (Brady)Rhizobium and legume communications will be an important approach likely to lead to solve the molecular basis of plant having "TOTIPOTENCY". © 1993 The Botanical Society of Japan.