Experiments were conducted during the winter seasons of 1983-1984 and 1984-1985 to identify suitable irrigation regimes s for wheat grown after rice in soils with naturally fluctuating shallow water table (SWT) at a depth of 0.4 to 0.9 m and medium water table (MWT) at a depth of 0.8 to 1.3 m. Based on physiological stages, the crop was subjected to six irrigation regimes viz., rainfed (Io); irrigation only at crown root initiation (I~); at only crown root initiation and milk (I2); at crown root initiation, maximum tillering and milk (I3); at crown root initiation, maximum tillering, flowering and milk (I4); and at crown root initiation, maximum tillering, flowering milk and dough (I5), Tube-well water with an EC <0.4 dsm -j was used for irrigation. Based on 166 mm effective precipitation during the cropping season, 1983-1984 was designated as a wet year and 1984-1985 with 51 mm as a dry year. The change in profile soil water content AW (depletion) in the wet year was less (23%) under SWT and 10% under MWT as compared to the dry year. The ground water contribution (GWC) to evapotranspiration (ET) was 58% under SWT and 42% under MWT conditions in both the years. The GWC in the wet year was 20% under SWT and 23% under MWT. Of the total net water use (NWU), about 85% was ET and 15% drainage losses. The NWU was highest (641 and 586 ram) in I s under SWT and MWT conditions, respectively, but not the yield (5069 kgha-l). Compared to 15, NWU in 12 treatment decreased by 10% in the wet and 25% in the dry year. A similar trend was observed in the 13 treatment under MWT condition. However, there was no statistically significant difference between yields of the I j to 15 treatments of either water table depth during the wet year. This was also true during the dry year for the I 2 to I s treatments. Under SWT, in 12, the grain yield was 5130 kgha -1 and under 13 regime, 5200 kgha -1. Under MWT in 13, the yield was 5188 kg ha -j and under 14 regime, 5218 kg ha i. Thus it appears that in the Tarai region