The systems productivity, soil fertility and N use efficiency were evaluated in a drought area of Rajshahi under five N fertilizer levels (0, 40, 80, 100 and 120 % N of the recommended dose, two straw retention (SR) (0 and 30%) and two tillage options [raised bed and conventional tillage (CT)] in a long term bed planting experiment with Rice-Wheat (RW) systems. The findings revealed that the permanent raised beds (PRB) with 30% straw retention had the highest productivity for all the three crops in the sequence. Within each N rate the total system (rice-wheat-mungbean) productivity was higher with 30% SR on PRB and the least in CT with 0 % SR. At 80 % of recommended fertilizer N rate, mean annual system productivity was 12.8 t ha-1 for PRB with 30% SR, 11.2 t ha-1 with PRB on 0% SR and 10.3 t ha-1 with CT without straw. N uptake and use efficiency increased with increasing N levels with bed planting up to 120% N application (120 kg N ha-1) in wheat, both 100% (80 kg N ha-1) in rice and (20 kgN ha-1 ) in mungbean for all the years. System productivity in N unfertilized plots increased when straw was retained. The results suggest that N fertilizer rates can be reduced when straw is retained. Soil organic matter in surface soil layers of the PRB had increased by 0.72% after eight years (8 rice-wheat-mungbean crop cycles) with 30% SR. It may be inferred that straw retention is an important component of soil management and may have long term positive impacts on soil quality compared with conventional tillage with 0 % SR. The combination of PRB with nutrients and residues retained appeared to be a very promising technology for sustainable intensification of RW systems in the drought prone area of Bangladesh.23-32