SUMMARYIn two experiments done in successive years to compare the growth and yield of Cappelle wheat either protected from or exposed to attacks by larvae of wheat bulb fly, the plots were previously fallowed, but egg laying was prevented on half of each plot by using Polythene soil covers. In the first season wheat was sown at the end of October, November and December and in the second season in late October and in early January, when there were two sowings, one with and one without a spring application of herbicide. The infestation rate was 1·1‐1·7 million eggs per acre, typical of a moderate attack. Larvae had little effect on the yield of October‐sown wheat as the plants had two shoots each when first attacked and few were killed. On plots sown late, yield was decreased by up to 22%, as plants had only a single shoot when attacked and many were killed.The main effect of wheat bulb fly was to reduce the number of ear‐bearing shoots by killing plants and restricting the production of new shoots. Surviving plants partially compensated by producing more ear‐bearing shoots with heavier ears and slightly heavier seeds than normal.Killed plants were not distributed uniformly but were often in patches several feet across. Wheat on the attacked plots ripened more slowly and unevenly than on the unattacked plots.Weather affects the growth of the plants and activity of the larvae and thus partly determines percentage shoot