Management of nitrogen (N) in rice-fish culture is of great significance in order to exploit synergies between rice and fish. However, detailed and systematic studies on N cycling in rice-fish systems are rare. An experiment was carried out at the Bangladesh Agricultural University using the fish species common carp, Cyprinus carpio L., and Nile tilapia, Oreochromis niloticus (L.). Efficiencies of fertilizer and feed as N inputs were evaluated by establishing an apparent N balance. The different N input strategies for rice-fish (RF) treatments were:(1) with fish, with urea (101 kg N ha À1 , RF Urea), (2) with fish, with feed (22.6 kg N ha À1 , RF Feed I), and (3) with fish, with feed (41.5 kg N ha À1 , RF Feed I I ) . R i c e o n l y ( v a r . B R 2 8 ) w i t h u r e a (101.2 kg N ha À1 ) was added as the control (Rice only). Urea as N input (Rice only, RF Urea) resulted in negative apparent N balances (N output-N input), whereas N input in the form of feed (RF Feed I, RF Feed II) led to positive apparent N balances due to the lower N input. The presence of fish increased the grain and straw N output in RF Urea compared to Rice only, the effect being significant for straw (P < 0.05). Moreover, a significant difference in nitrogen uptake by rice was seen between the two feeding levels. The higher feeding rate in the RF Feed II treatment led to higher N outputs of grain, straw and fish compared to the RF Feed I treatment, a significant difference occurring for straw (P < 0.05). An apparent recovery efficiency of 48.7% was determined for the additional N input in RF Feed II compared to the RF Feed I teatment. An economic analysis revealed that the RF Urea treatment was the most profitable, as reflected by the highest gross margin. In conclusion, the results indicate that fish can improve nitrogen uptake by rice plants. Moreover, the supply of supplementary feed has a fertilizing effect on the rice due to the utilization of excreted feed nutrients by rice plants.