The nature of the soil, the influence of previous natural grassland, ley, fallow or cultivation, and the weather conditions before and after planting are shown to affect the yield of wheat, and its response to fertilizers, especially nitrogen. Out of seventeen experiments over four seasons, involving nitrogen top-dressings on wheat, thirteen gave economic increases after the application of I cwt per acre of ammonium sulphate nitrate (A.S.N.). Such applications to the first, second or third wheat crops after Rhodes grass (Chloris gayana) ley, or to land continuously cropped for a number of years, nearly always gave an economic increase in grain yield, but on continuously cultivated land, with normal rainfall after planting, an economic return to a second cwt per acre was probable. Following a one-season fallow, economic responses to nitrogen only occurred when the pre-planting period was very wet. Yields of wheat after ley, both with and without applied nitrogen, tended to be superior to those on continuously-cropped land. Responses to phosphorus only occurred in one area, where the importance of applying both N and P, particularly to newly-ploughed grassland, was very evident. Loss of organic matter, and structural deterioration accompanied by lower efficiency in the utilization of water and nitrogen, are probably the main causes of falling yields on soils cultivated for many years. The use of grass/legume leys is suggested as the best means of maintaining these naturally fertile soils in a highly productive state, even though the present returns from the legs themselves may be less profitable than wheat cropping.