Ontario soft white winter wheat naturally contaminated with deoxynivalenol (DON, vomitoxin) at 0.45 mg/kg was cleaned and milled in industrial, pilot, and experimental mills. Some of the industrially milled fractions were baked commercially into a variety of food products. The samples were analyzed for DON and, in some instances, the fungal metabolite ergosterol. Milling led to a fractionation of both DON and ergosterol, with increased levels in the outer kernel (e.g., bran) portions and decreased levels in the inner flour portions. The effect of baking on non-yeast products was variable, ranging from no effect to 35% reduction, and when combined with milling and dilution effects gave an overall reduction of 66 ± 20% in DON. Experimental milling of two samples of naturally contaminated (at 1.2 and 6.9 mg/kg) Quebec hard red spring wheat resulted in a similar fractionation of DON. A positive correlation between DON and ergosterol levels in 11 eastern Canadian wheat samples indicated that the level of DON production was directly related to the incidence of fungal growth; the milling data suggested that fungal infection was greatest at or near the kernel surface.4-Deoxynivalenol (DON, vomitoxin, 3a,7a,15-trihydroxy-12,13-epoxytrichothec-9-en-8-one), a naturally occurring metabolite produced by the fungus Fusarium graminearum Schwabe on a variety of cereal grains, is a member of the trichothecene class of mycotoxins known to be associated with several diseases in animals and man (Ueno, 1980).Following the discovery of DON at levels up to 8.5 mg/kg in some of the 1980 soft white winter wheat crops in Ontario (Scott et al., 1981;Trenholm et al., 1981), the Canadian Government set a maximum guideline level of 0.3 mg/kg in the finished product for that crop destined for use in nonstaple foods. DON also appeared in the 1981 and 1982 crops (averaging 0.7 mg/kg in 1982) as well as occurring in the 1980 and 1981 Quebec hard red spring wheat crops. Soft winter wheat is typically used in nonstaple foods such as biscuits, cakes, cookies, and bran breakfast cereals, whereas hard spring wheat is used in staple foods such as bread and related products.Little was known at the outset of this study concerning the fate of trichothecenes in general and DON in particular during grain processing. Collins and Rosen (1981) reported that wet milling of corn removed about two-thirds of the T-2 toxin present, and Kamimura et al. (1979) noted the effects of processing on a variety of oriental foods containing trichothecenes. Recent studies (Scott et al., 1983;Hart and Braselton, 1983) showed that milling of hard spring wheats had little palliative effect and that DON was distributed throughout the milled products. Baking the flour into bread failed to destroy the DON (Scott et al., 1983;El-Banna et al., 1983).This paper reports the results of studies to determine the fate of DON during wheat processing. The distributions of DON in milling fractions produced from commercial, pilot, and experimental milling of soft wheats were