This paper seeks to show that cereal breeders have been successful when their objectives have been clear. It is suggested that the end of cereal improvement is not in sight and that breeders will continue to make significant advances in many directions. In order to look at future trends it is necessary to examine what has been achieved and to consider the principles on which cereal breeding is based.As crop improvement is a consciously directed form of evolution, it is worth remembering that all plant species, whether cultivated or not, have evolved from species and genera quite different from themselves. The capacity for change is enormous. Cereal crops have had the course of their evolution modified by man's activities since before recorded history, but only during this century have deliberate attempts been made to improve the performance of cereals based on a knowledge of inheritance. Indeed the scale of cereal breeding remained quite small throughout the first half of this century. The considerable advances achieved should, therefore, be measured against the comparatively short history of scientific breeding.Observations of varieties in cultivation in the first half of this century show that their straw heights and life spans have both been reduced compared with more recent varieties. Furthermore, varieties released in the last decade have shown a continuing height reduction, but not at the expense of yield.Increases in grain yield are often attributed to better husbandry, more fertilizer and easier harvests, but frequently it is the improvement in the varieties, for example stiffer straw and earlier ripening, which allows the increased yields to be secured. The effect of large seasonal fluctuations can be reduced by plotting the 5-year national means (Agricultural Statistics, H.M. Stationery Office). These are shown separately in Fig. I for wheat, oats and barley in Scotland and in England and Wales for the periods 1945-49 until 1972-76. With few exceptions the trends for all the curves have been continuously upwards, with the Scottish figures higher than the southern ones for barley and wheat but not for oats. There is no evidence that a limit to yield has been reached.Improvements in quality characters have often preceded an understanding of their biology and genetics. Long before much was known about the biochemistry of these attributes breeders wem successfully selecting for them in an empirical way using correlated visual characters. It is still true that the biochemistry of good malting quality is being analysed and as each new component is described it becomes possible to define the objectives in more precise terms. When the barley at https://www.cambridge.org/core/terms. https://doi