Determinations have been made of the amounts of HCN in bracken at various stages of maturity and of the apparent retention of cyanide by animals fed on fresh bracken, dried bracken and hay plus sodium cyanide. Newly emerging bracken shoots contained up to 50 mg. HCN per 100 g. dry matter but the mean value for early‐summer bracken (18 June to 16 July) was only 3·35 mg. per 100 g. dry matter and that for early‐autumn plants (18 Aug. to 27 Sept.) only 1·81 mg. per 100 g. A young second‐growth in October contained an average of 7·35 mg. per 100 g. There were considerable day‐to‐day fluctuations in the HCN content of bracken. As it was difficult to arrive at valid figures for normal urinary cyanide and thiocyanate no attempt was made to correct the urinary excretion values obtained for bracken‐fed animals but the apparent retention of HCN was nevertheless high in nearly all cases. HCN destroyed in the digestive tract or excreted through the lungs could not be measured, these fractions being included in the figures for apparent retention; the high apparent retention of HCN by bracken‐fed animals is therefore not regarded as sufficient evidence to indicate the cyanogenetic principle as a causative factor in ruminant bracken ‘poisoning.’ The relatively low HCN content of bracken during most of the season, the insignificant amounts of cyanide found in the liver and rumen contents of bracken ‘poisoned’ animals and the failure of thiosulphate injections to promote recovery suggest that danger arising from the cyanogenetic principle of bracken is likely to be slight. The administration of sodium cyanide did not lead to any increase in the combined urinary excretion of cyanide and thiocyanate, indicating a difference from the metabolism of the cyanogenetic principle of bracken.