Abstract:One major reason legumes are utilized in farming systems is for their ability to maintain the soil nitrogen balance when incorporated into crop rotations. As such legumes can play an important role in both maintaining and improving crop yields. It is therefore important to quantify the total amount of nitrogen fixated through the biological nitrogen fixation (BNF) process used by legumes and to determine how much of this fixated N is available to the following crop. In doing so, this will enable greater precision in terms of appropriate fertilizer application rates and can increase fertilizer use efficiency. In 2002 and 2003, the 15N natural abundance technique was used to record the percentage of biological nitrogen fixation (%Ndfa) fixated by soybeans and the soybeans N contribution to a spring maize-winter wheat-summer soybean rotation system. The field experiment which was located in the Gansu Loess Plateau used four tillage treatments. These were: conventional tillage (T), conventional tillage with stubble retention (TS), no-tillage (NT), and no-tillage with stubble retention (NTS). In general the %Ndfa by soybean was higher in 2003 than in 2002, related to 2003 receiving a significantly more rainfall. The %Ndfa by soybean in the TS treatment was significantly higher than in the other three treatments. The TS treatment benefited f rom a higher N fixing enzyme activity which improved the soybeans BNF ability and as a consequence resulted in a higher yield. A negative relationship between the amount of N fixed by soybean and the amount of NO3-N in the 0-30cm soil profiles was found. A soil NO3-N concentration of 40 kgN/hm2 was thought to be the critical value above which BNF activity was reduced. Soybean obtained its N requirements from both soil N and BNF. When there was no BNF activity recorded, N was supplied wholly by soil N, which led to soil N depletion and soil N scarcity. The aboveground biomass of soybeans increased with increases in %Ndfa, however this finding was most evident in the TS treatment. A positive relationship between the amount of N fixed by soybean and its above-ground biomass was also found. In 2002, the treatment effects on the proportion of soybean BNF to total N input ranged from 13~22%, with the highest proportion recorded under the TS treatment. In 2003, the %Ndfa ranged from 43.8 %~ 62.4%, with the amount of fixated N ranging from 10.7~17.6kgN/hm2 and the proportion total N contributed by N fixation ranging from 40%~63%.