Increasing crop N use efficiency and minimizing environmental risk require an accurate assessment of N taken up by the crop from different sources. We conducted this study to: (i) compare the grain yields of corn (Zea mays L.) in monoculture and in rotation with soybean [Glycine max (L.) Merr.]; (ii) determine the contributions of N from fertilizer, soil, and legume residue to corn in the rotation; and (iii) compare N fertilizer recovery in monoculture and in rotation. Two existing (>10 yr) irrigated corn-soybean rotation areas in Kansas were used. The soils were Crete silt loam (fine, smectitic, mesic Pachic Argiustolls) and Eudora loam (coarse-silty, mixed, superactive, mesic Fluventic Hapludolls). To trace the N through the rotation, 15 N microplots (2.4 m 2 ) were established in the corn. Microplots also were established in soybean to separately follow 15 N from roots + soil and shoots to corn. Crop rotation and fertilizer addition increased corn yield at both sites for two growing seasons. Averaged for 2 yr, the amount of N needed in the continuous corn to achieve yield equal to that in rotation with no N added was equivalent to 144 kg N ha -1 in the Crete silt loam and 155 kg N ha 'in the Eudora loam. Response to N was greater on the Eudora loam, probahly because of textural and organic matter differences, In the Eudora soil, significantly higher amounts of soil N were taken up at harvest by corn in rotation, whereas, in the Crete soil, corn in monoculture took up significantly higher amounts of soil N. Corn plants recovered 3 kg N ha 1 (3%) from soybean residue in the Eudora soil and 5 kg N ha -1 (14%) in the Crete soil. The main value of legume residue appears to be long-term maintenance of soil N to ensure adequate delivery to future crops.