There is continuing concern about the role of irritation in cancer development. Methyl bromide, a widely used fumigant and known irritant reported to cause forestomach carcinomas in rats, was dissolved in peanut oil and given by gavage at 50 mg/kg body wt to Wistar rats five times per week for 13 to 25 weeks. Starting at Week 13, methyl bromide administration was discontinued for half of the methyl bromide-treated rats (stop treatment group). After that, rats from both the continuous treatment and stop treatment groups were terminated at 4-week intervals to follow the progression of the stomach lesions. Forestomach lesions were not found in control rats receiving peanut oil and killed at 13 or 25 weeks. At 13 weeks the forestomachs from rats receiving methyl bromide were contracted and adherent to the liver and spleen. Inflammation, acanthosis, fibrosis, and a high incidence of pseudoepitheliomatous hyperplasia were found microscopically in treated animals. At 25 weeks, 100% of the rats receiving methyl bromide continuously had hyperplastic lesions of the forestomach which were more severe than those at 13 weeks. Evidence of malignancy was seen in one rat and the lesion was considered a very early carcinoma. In the stop treatment group that received methyl bromide for 13 weeks, there was regression of the stomach lesions, but at the 12-week final sacrifice, adhesions, fibrosis, and mild acanthosis remained. This study illustrates the need for regression experiments for complex forestomach lesions in rodents, especially when an irritating chemical is given by gavage.