Yield of a dryland crop such as wheat is affected by environmental, management and genotypic factors. Field experiments and computer simulations were conducted to understand the effect of supplemental irrigation, type of cultivar and time of sowing (using simulation) on grain yield and quality of two wheat varieties EGA Gregory and Livingston. There was significant effect of variety and watering regime on grain yield, harvest index, kernel weight and number and protein content. The maximum grain yield of 6.10tha-1 was obtained from the irrigated mid-late flowering variety EGA Gregory in the year the crop was sown early. The effect of irrigation on the late flowering variety EGA Gregory was higher than that on the early flowering variety Livingston. The grain water use efficiency of EGA Gregory was higher than that of Livingston in the year it was sown early. APSIM simulations of yield using historical weather data of the past 114 years show that as the sowing date is delayed, the decrease in yield of EGA Gregory becomes higher than that of Livingston and by mid-June sowing, the grain yield of Livingston becomes higher than that of EGA Gregory. As the sowing is delayed, the differences between the anthesis dates, and also grain yields of the two varieties get smaller. The simulation results show that the grain yield of late-sown EGA Gregory and Livingston increases with the amount of in-crop rainfall. However, when sown early, the yield of Livingston decreases with the amount of in-crop rainfall. Farmers need to have different cultivars on hand and sow according to seasonal condition. An early-sown mid-lateflowering cultivar is less risky.