Butylated hydroxyanisole (BHA) induces tumours of the squamous epithelium of the forestomach of rodents, but not at other sites. Although humans do not have squamous epithelium in their stomach the likelihood that BHA will induce tumours of the squamous epithelium of the oesophagus needs to be considered. Studies in several species indicate that the forestomach epithelium is very responsive to hyperplastic and neoplastic change induced by BHA, but the oesophageal epithelium is not responsive. The lack of effect in the oesophagus is likely to be due to the fact that the rapid speed of transit through the oesophagus limits the exposure time of the oesophageal mucosa to the food contents. Conversely, as the rodent's forestomach has storage function, exposure of the squamous epithelium of the forestomach would be continuous. The fact that the no-observed-effect level for hyperplasia of the oesophageal mucosa is several hundred times the acceptable daily intake for BHA supports the view that BHA would not be a human carcinogen at food additive levels of use.