The carcinogenicity of β-cyclodextrin, a cyclic, water-soluble carbohydrate comprising seven glucose units, was examined in Fischer 344 (F344) rats. Groups of 50 males and 50 females were given the compound in their diet at concentrations of 0 (control), 2.5 or 5% for 104 wk. Surviving rats were then given a basal diet for a further 5 wk and killed at 109 wk. The dose levels were selected from the results of a 13-wk subchronic toxicity study. Dose-dependent inhibitory effects of β-cyclodextrin on growth were observed in both sexes of the treated groups. The survival rates, mean survival times and range, however, demonstrated no significant differences between the control and treated groups. A variety of tumours developed in all groups, including the control group, but all the neoplastic lesions were histologically similar to those known to occur spontaneously in this strain of rat, and no statistically significant increase in the incidence of any tumour was found for either sex of the treated groups. Thus, it is concluded that under the present experimental conditions, the high dose, about 340–400 times higher than the current daily human intake from ingestion as a food additive and from pharmaceutical use, does not have any carcinogenic potential in F344 rats.