Field experiments were conducted at the Crop Research Centre of Govind Ballabh Pant University of Agriculture and Technology, Pantnagar during 1996/97 and 1997/98. Each experiment comprised 10 crop sequences: (a) wheat–rice, (b) chickpea–rice, (c) lentil–rice, (d) pea–rice, (e) wheat–mungbean green manure–rice, (f) wheat–Sesbania green manure–rice, (g) wheat–fodder–rice, (h) chickpea–fodder–rice, (i) lentil–fodder–rice and (j) pea–fodder–rice, in a randomized block design with four replications. The crop sequences were compared in terms of economic rice equivalent yield (REY), protein production, apparent nutrient balances and effect on soil fertility status. Amongst crop sequences involving two crops each year (200% cropping intensity), chickpea–rice gave highest REY and protein production. Of the sequences involving three crops each year (300% cropping intensity), chickpea–fodder–rice and wheat–fodder–rice were superior to others. The P balances were positive for all sequences, whereas K balances were generally negative except for sequences involving green manure legumes. Green manuring with Sesbania or mungbean helped restore soil fertility, indicating the advantage of green manure for higher productivity and sustainability of rice–wheat system. Chickpea–rice and chickpea–fodder–rice appeared promising alternatives to rice–wheat crop sequence.