Summary Bracken poisoning has been produced experimentally in sheep by feeding fresh bracken, although it was found difficult to induce sheep to eat adequate quantities of the plant. The apparent absence of natural cases in sheep is probably due to their selective grazing habits. Bracken poisoning in sheep produced the same symptoms and post-mortem findings as in cattle. The poisoning of ruminants fed on bracken appeared to be closely related to the stage of growth, young bracken being most toxic. Bracken poisoning was also produced in sheep by feeding artificially dried bracken, although a greater consumption of dry matter was found to be necessary when this material was used. There were indications that the drying process may reduce the degree of toxicity of the plant. Whole bracken cubes did not show any poisonous effect when fed to sheep at a relatively high level (1·38 Kg. per day) for 44 days. Cubed alcohol-extracted bracken fed to a sheep for a similar time and in even larger amounts (1·95 Kg. per day) had no poisonous effect and was quite palatable. The live weights of sheep fed entirely on bracken were closely related to the quantities of dry matter consumed. When sufficient bracken (fresh, dried or cubed) was consumed its nutritive value was clearly demonstrated by the gains made.