Growth and yield of wheat are affected by environmental conditions and can be regulated by sowing time and seeding rate. In this study, three sowing times [winter sowing (first week of September), freezing sowing (last week of October) and spring sowing (last week of April)] at seven seeding rates (325, 375, 425, 475, 525, 575 and 625 seeds m(-2)) were investigated during the 2002-03 and 2003-04 seasons, in Erzurum (Turkey) dryland conditions, using Kirik facultative wheat. A split-plot design was used, with sowing times as main plots and seeding rates randomized as subplots. There was a significant year X sowing time interaction for grain yield and kernels per spike. Winter-sown wheat produced a significantly higher leaf area index, leaf area duration, spikes per square metre, kernel weight and grain yield than freezing- and spring-sown wheat. The optimum time of sowing was winter for the facultative cv. Kirik. Grain yields at freezing and spring sowing were low, which was largely the result of hastened crop development and high temperatures during and after anthesis. Increasing seeding rate up to 525 seeds m(-2) increased the spikes per square metre at harvest, resulting in increased grain yield. Seeding rate, however, was not as important as sowing time in maximizing grain yield. Changes in spikes per square metre were the major contributors to the grain-yield differences observed among sowing times and seeding rates. Yield increases from higher seeding rates were greater at freezing and spring sowing. We recommended that a seeding rate of 525 seeds m(-2) be chosen for winter sowing, and 575 seeds m(-2) for freezing and spring sowing.