A 3-year study was conducted to measure the effect of sowing time and seeding rate on the development and yield of wheat (Triticum aestivum L.) grown under high-yielding conditions in north-eastern Victoria. A range of wheat cultivars with different development responses, including 'winter' types, was used in 2 experiments in each season. High grain yields for dryland wheat were measured in the first 2 seasons (1985-86), and in 1985, near-optimal water use efficiencies (>18 kg/ha. mm effective rainfall) were obtained. In the third season (1987) grain yield was limited by adverse climatic conditions-in the me- and post-anthesis period. In each season, grain yield declined with delay in sowing time. In 1985 there was a loss of 200-250 kg grain/ha for each week's delay in sowing time. In 1987, yield loss with delayed sowing was 50-110 kg grain/ha. In each season, cultivars with late or midseason maturity development gave the highest mean yields, and the use of these maturity types allowed earlier sowing, in mid April. However, with late sowing of wheat there was a trend for early maturity types to give higher yields, and so the use of 2 wheat cultivars with distinct maturity development responses to climate is recommended. If only 1 wheat cultivar is to be used, then a late maturity type is recommended. Higher wheat yields were also obtained as spike density increased, as a result of higher seeding rates. Our data suggest that in the higher rainfall region of north-eastern Victoria, a spike density of about 500 spikes/m2 is required to optimise wheat yields.