Seven species or varieties of grass, and a mixture of 3 of them, were sown in pure swards, treated with 4 levels of nitro‐chalk (0, 17·5, 35, and 70 Ib N/ac/cut) and cut 4 or 5 times each year. Each species and the mixture were also sown with white clover, and the effect of fertilizer N on the yield of N in each grass was compared with the effect of clover on the yield of N harvested from the grass/clover swards.The regression line for response in yield of N with increasing levels of fertilizer N showed slight, but significant, upward curvature. The grasses differed in their uptake of N from the soil, S37 cocksfoot and S48 timothy showing relatively high uptakes, and the ability of each grass to take up fertilizer N was usually related to its uptake of N from the soil. Differences in the yield of dry matter between the species, at a similar level of N, are discussed, and it was concluded that perennial ryegrasses were most efficient andAgrostis tenuiswas least efficient in using the N taken up in the production of DM.In general, there were no significant differences between the yields of N of the grass/clover mixtures; the N yields of the grass components were significantly different and tended to be inversely related to the N yields of clover.Grasses which gave high yields of N with fertilizer were also high yielding when grown in association with clover. Pure grass swards required more than 200 Ib fertilizer N/ac/yr in order to yield the same amount of N as the grass/clover swards. The amount of N estimated to have been derived by grass from clover (indirect effect of clover) increased each year; it was highest with S37 cocksfoot and lowest with Irish perennial ryegrass, averaging 46 and 23 Ib N/ac/yr, respective