The relationship between deoxynivalenol (DON) intake and first morning urinary DON was examined in UK adults to validate the latter as a biomarker of human exposure. DON was assessed in first morning samples collected during a period of normal diet, a wheat-restriction intervention diet, and partial wheat-restriction intervention in which bread was allowed. During the partial intervention duplicate bread portions were collected for DON analysis. During the normal diet, partial intervention and full intervention, urinary DON was detected in 198/210 (geometric mean 10.1 ng DON mg(-1) creatinine, 95% confidence interval (CI) 8.6-11.6 ng mg(-1); range nd-70.7 ng mg(-1)), in 94/98 (5.9 ng mg(-1), 95% CI 4.8-7.0 ng mg(-1); range nd-28.4 ng mg(-1)), and 17/40 (0.5 ng mg(-1), 95% CI 0.3-0.7 ng mg(-1); range nd-3.3 ng mg(-1)) volunteers, respectively. A strong correlation between DON intake and the urinary biomarker was observed (p <0.001, adjusted r(2) = 0.83) in models adjusting for age, sex and body mass index. These data demonstrate a quantitative correlation between DON exposure and urinary DON, and serve to validate the use of urinary DON as an exposure biomarker.