Lotus pedculatus L., Medicago sativa L., Macroptilwum atropwrpureum, Glycine max, and Trifoliun repens L. were grown in a N-free medium and inoculated with one of ten Rhizobium strains. Dry matter, N content, and 8"N values were determined for various plant parts.Nodules, with the exception of those from lucerne, were enriched in 1"N relative to atmospheric N. Considerable variation was found in 81"N values of plant herbage (-4.5 to +0.8). The extent of isotopic discrimination was dependent on both the host plant and the infecting rhizobial strain. This further complicates, but does not invalidate, the use of small variations in the natural abundance of 1"N to estimate the contribution of symbiotically fixed N2 to the N in legume herbage. Some other implications of the observed differences are also discussed.Legumes obtain their N requirements from two sources: symbiotic N2 fixation and assimilation of soil mineral N. Although a quantitative partitioning of both sources is important in understanding the N nutrition of legume crops in mono-culture or in mixed species stands, there are relatively few reports in the literature of field measurements of symbiotic N2 fixation, largely because of the difficulties of making such measurements (14).Many investigators have reported small but measurable differences in the natural abundance of "N in N2-fixing and nonfixing plants growing in association, and some have suggested that these differences may be used to provide a quantitative integrated measurement over time ofthe contribution ofN2 to the N economy of the fixing plant (2,5,13,16,20 (22). This raises the question as to whether the isotopic composition of the host plant is influenced by environmental parameters or by the Rhizobium strain with which it is infected.Nodules of some N2-fixing plants are elevated in "N compared to other plant tissues (18) and it has been suggested that "N enrichment is a result of the denitrifying capability of some rhizobia (20) although this has subsequently been rejected (10). A more recent hypothesis is that "N enrichment in nodules is indicative of the predominant type of N compound transported from the nodule (18). Inasmuch as "N enrichment in nodules will affect the "N concentration in harvested plant material, it will influence estimates of N2 fixation based on variations in the natural abundance of "5N.