The effect of four seed rates on the grain yield, grain N content and grain size of Blenheim malting barley sown on two dates in the spring was tested on three different soils over a 4-year period (1991–94). The four seed rates (120, 160, 200 and 240 kg/ha) gave mean plant populations of 222, 308, 374 and 430 plants/m2, respectively, for the first sowing date (February to early March) and 235, 340, 405 and 470 plants/m2 for the second sowing date, 3–5 weeks later (late March to mid-April). The earlier-sown crops generally gave greater yield, lower grain N and smaller amount of screenings, but in two of the twelve experiments, the first sowing gave lower yield and greater grain N.The influence of seed rate on yield and quality depended on sowing date. The three higher seed rates (160, 200, 240 kg/ha) had little influence on yield or grain N, irrespective of sowing date, and there was no significant difference in yield between the lowest (120 kg/ha) and the three higher seed rates when the crop was sown in February and early March. But when the crops were sown in April, the lowest seed rate (120 kg/ha) gave the lowest yield in all seven comparisons and gave significantly reduced yield in three of the seven comparisons. There was a significant increase in grain N in three of the comparisons. Seed rate had little effect on grain screenings or 1000-grain weight. Soil differences did not influence the effect of seed rate on the yield and quality of the grain.