Davies (1941) has estimated that there are nearly 1,300,000 acres of hill land in England and Wales upon which the white bent or wire grass (Nardus stricta) is predominant, or constitutes a high proportion of the ground cover. From this estimate, which is probably a conservative one, it follows that approximately 24% of the rough grazings of this country are wholly or largely Nardus pastures. The opinion that Nardus is of almost negligible grazing value is widely held. Smith (1918) has said that no stock will graze it by preference, although he admits that a certain amount of the foliage is eaten in early summer. Pearsall (1950), Fenton (1936) and Stapledon (1937) have expressed very similar views. Wallace (1917) regarded this grass in a more favourable light and, while he certainly did not rank it with the best moorland grazing plants, stressed its value in winter. The truth is that these opinions are based almost wholly on observation and are supported by very little evidence on nutritive value. Analyses on conventional lines have been made by Kinch (1884), Evans (1927), Fagan & Watkins (1934) and Thomas & Dougall (1948). Thomas & Trinder (1947) have made a detailed examination of the ash. Digestibility data appear to be nonexistent; without such data any worth-while estimate of nutritive value is almost impossible. In view of the existence of very large areas on which little other than Nardus is available for the sustenance of hill sheep, it would seem desirable to attempt such an estimate. It was therefore decided to carry out digestibility trials in which the grass, cut at two stages of growth, was fed to sheep of a hill breed.