Estimates of the contribution of biologically fixed N to the total N of nodulating soybeans (Glycine max (L) Merrill, variety Harosoy) grown under a variety of conditions were made from: (a) differences in N yield between nodulating and nonnodulating isolines; and (b) differences in 15N abundance between the two isolines. For plants grown in a greenhouse in nutrient-poor soil, both estimates showed a high level of N2 fixation; from 58 to 89% N fixed by differences in N yield and from 51 to 95% by differences in "5N abundance. Decreasing contributions of fixed N were estimated by both methods with increasing levels of added N03-. Results of field experiments carried out over two years on an unamended highly fertile midwestern soil showed a modest level of N2 fixation by both methods (7.3 to 51% by differences in N yield, and 5.4 to 46% by differences in 15N abundance). When the soil was amended with ground corn cobs, both methods showed higher contributions of fixed N. The two methods of estimating N2 fixation gave similar results. Both appear to be semiquantitative and the standard errors of the estimates were about the same (6% on the average).A number of investigators (1-4, 9, 10, 11, 14) have reported small differences in the 15N abundance between N2-fixing and non-N2-fixing plants, with the 15N abundance in N2-fixing plants being consistently lower than in nonfixing plants. These investigators have suggested that this difference could be used to indicate the contribution of atmospheric N2 to the total N of the N2-fixing plants. Such a method requires that: (a) there be a measurable difference in the "5N abundance between the two major sources of N to N2 fixing plants (atmospheric N2 and soil-derived N); and (b) this difference be maintained after alteration of 15N abundance resulting from isotopic discrimination on uptake of N. If portions of the plant, rather than the entire plant, are sampled, the method also requires that isotopic discrimination associated with metabolic processes, and subsequent transport of isotopically altered N compounds, does not result in large distortions of the original "5N abundance in the sources of the N.In applying the natural "5N abundance method, the 15N abundance in non-N2-fixing plants is used as a measure of the 15N content of soil-derived N, after it has undergone any fractionation that may occur during N transformations in the soil and on uptake into the plant. The 1 N abundance of the N2-fixing plant should