A field experiment was conducted during, pre-rainy season, rainy season, winter season and summer season of 1989-91 to evaluate the productivity, fertility build-up of soil and the economics involved in different rice (Oryza saliva L.) - based cropping systems. The productivity of rice was maximum (4.98 and 5.69 tonnes/ha) when both rainy- and summer-season rice in sequence were green-manured in situ with prickly sesban or dhaincha [Sesbania cannabina (Retz.) Pers.; syn S. aculeata (Willd.) Poir.], grasspea or khesari (Lathyrus sativus L.) respectively along with 100% of the recommended dose of N, P and K (60, 14.2 and 25 kg/ha and 80, 22 and 41.5 kg/ha respectively in rainy- and summer-season rice). However, comparable yield (4.84 and 5.42 tonnes/ha) was obtained when both rainy and summer-season rice in sequence were green-manured in situ along with 75% of the recommended dose of fertilizers (N45P10.6K18.7 and N-60 P16.5K31 in rainy and summer season respectively). Green-manuring twice in a year before transplanting of both rainy- and summer-season rice improved the soil-nutrient status (260 kg N/ha, 16.1 kg P/ha and 21 kg K/ha). Green-manuring with sesban and khesari was found equivalent to 19.4 and 31.1 kg fertilizer N/ha during rainy- and summer-season rice respectively. Maximum net production value (2.11) was recorded in the sequence where both the rice crops were green-manured with legume along with 75% recommended dose of N, P and K. Fertilizer-use efficiency was improved (40.05)when any legume crop was included in a cropping sequence involving 2 rice crops, and the improvement was maximum (41.42) when legume crop was buried into soil before transplanting both rainy- and summer-season rice. Hence the recommended dose of N, P and K may safely be reduced by 25% if green-manure is provided to both the rice crops in sequence.