The average yields of a 3-year field experiment (1985/86–1987/88) on a mollisol at Pantnagar showed that, at 87 kg N/ha, applying urea supergranules (USG) gave 0·47 t/ha more rice grain than split application of prilled urea (PU). The data also showed that, at this rate of N, one-third (29 kg N/ha) of inorganic N can be substituted by fresh straw (wheat or rice) applied at planting to give similar rice yields. However, at a total of 58 kg N/ha, 29 kg N/ha applied as straw with 29 kg N/ha of PU or USG reduced yield compared with 58 kg N/ha of PU or USG alone, except in 1986/87. Yield differences were attributed to differences in number of panicles/m2, N uptake by the crop and mean wet soil NH4+-N obtained at various stages of crop growth during 1986/87 and 1987/88. Wheat grown after rice in 1985/86 and 1986/87 showed significant residual effects in plots with straw application and these were correlated with the rate of hydrolysable organic N (HN) obtained after rice.