A two-year field experiment on a mollisol at Pantnagar showed that organic sources of N gave generally higher rice yields than the equivalent split dressing of prilled urea (PU). In 1985/86, Sesbania green manuring (40 kg N/ha) and PU (40 kg N/ha) and, in 1986/87, Sesbania (40 kg N/ha) + Azolla (40 kg N/ha) gave significantly higher yields than a split application of 80 kg N/ha as PU. Yield increases were attributed to significant variations in panicles/m2, filled spikelets/panicle and nitrogen uptake by the crop. These components, and grain yield, were significantly correlated with wet soil NH+4-N, which was increased by organic manuring. Wheat grown after each rice crop showed significant residual effects of the different sources of N applied to rice.