A worker provided a routine faecal sample for plutonium and americium analysis. In the course of this analysis 500 mBq of 228Th was discovered. There seemed no credible occupational route for intake of thorium. Further investigation revealed that the worker consumed ∼25 g d-1 of nuts, including Brazil nuts. A sample of these nuts was analysed and found to contain activities of 228Th in sufficient quantity to account for the faecal activity. However, follow-up urine samples taken from the worker showed 0.6-0.7 mBq of 228Th. The intake of 228Th via nuts is insufficient to account for this activity in urine. However, it is likely that the intake of 228Th was accompanied by similar activity of the parent 228Ra, and biokinetic calculations show that decay of 228Ra in vivo would produce sufficient 228Th to account for the observed urine activity.