Field experiments were conducted at Tamil Nadu Rice Research Institute, Aduthurai, India, during the wet seasons of 1992 and 1993 to study the effect of full and partial substitution of fertiliser N with green manure N (Sesbania rostrata) on nitrogen uptake, yield attributes and yield of rice. The experiment consisted of eight treatments with two levels of N (100 and 200 kg ha−1) and three sources of N application viz., fertilizer, integrated (1:1 fertilizer and green manure N) and green manure N compared to the recommended practice (150 kg fertilizer plus 6.25 t ha−1 (72 kg N) green manure) and a no N control. Nitrogen application markedly increased the N uptake. Combined use of the two N sources at 200 and 222 kg N ha−1 and of single fertilizer N at 200 kg N ha−1 recorded the maximum N uptake, increased the yield attributes such as number of panicles per unit area, weight per panicle, number of total and filled grains per panicle and test weight. At 200 kg N ha−1 full substitution of N by green manure reduced the grain yield but only partial substitution of N by green manure resulted in almost similar yield as single fertilizer N. Thus 200 kg N ha−1 applied in equal proportions of fertilizer and green manure N can be recommended for medium duration rice cultivars.