The variable responses of crops to added nitrogen (N) in Alfisols of the Indian semi-arid tropics are partly due to variable rainfall and partly due to variable losses of available-N. To measure the losses of N through leaching, which can be appreciable under some circumstances, a field experiment was conducted during the rainy season (June-September) of 1992, using bromide (Br) as a tracer for NO 3. Bromide (as NaBr) was applied to bare fallow soil at a rate of 200 kg ha-I in microplots (2 m × 2 m) and its vertical movement was monitored periodically. Data on rainfall and Br-distribution in the soil profile on different dates of soil sampling clearly indicated that the movement of Br-was strongly dependent on rainfall. During the first month (15 June-15 July) after Brapplication, with scattered and light rainfall about 90% of the added Br-remained in the soil profile (0.6 m). After continuous heavy rainfall in early August more than 90% Br-had moved beyond 0.6 m depth. This indicates a very high risk of NO 3 leaching in this soil, and it is unavoidable without special measures to protect the applied N.