SINCE our initial reports of the carcinogenic potential of bracken for rats and mice1,2 part of our work has been directed towards the isolation of the causative factor(s). The assays used have included acute toxicity testing in mice (orally and intraperitoneally (i.p.)) and mutagen trials with Drosophila and mice, as well as long term cancer induction in rats and mice. The former suffer from the disadvantage of being nonspecific for carcinogens and the latter from the delay of a year or 15 months before results can be assessed. A fraction was isolated, however, which proved positive for acute toxicity, mutagenicity and Carcinogenicity in mice and the preliminary molecular formula of C7H8O4 was given for the main constituent3,4.