An experiment to study the effect of sowing time (21 October, 5 November, 20 November, 5 December and 20 December), seed rate (5, 7·5 and 10 kg/ha) and application of nitrogen fertilizer (0, 20, 40 and 60 kg N/ha) was conducted at the Punjab Agricultural University, Ludhiana, on a loamy sand soil during the winter seasons of 1974–5 and 1975–6. The delay in sowing from 21 October to 20 December produced shorter plants with fewer spikes. The crop sown on 21 October using 7·5 or 10 kg seed/ha and supplied with 20 or 40 kg N/ha gave better yield than those of later sowings.