Background: During 1998, weather conditions in the United States favored the growth of Aspergillus species leading to widespread contamination of Midwestern and Southern corn with hepatotoxic and hepatocarcinogenic aflatoxins. We designed a study to provide the first national prevalence estimate of aflatoxin exposure using the National Health and Nutrition Examination Survey (NHANES), a representative cross-sectional survey of the noninstitutionalized civilian population of the US. Methods: Isotope dilution liquid chromatography-tandem mass spectrometry was used to quantitate serum concentrations of aflatoxin B1-lysine in a one-third random subset of participants from NHANES 1999-2000. Results: About 1% of the U.S. population had detectable levels (³ 0.02 _g/l) of aflatoxin B1-lysine. Of those with detectable levels, the geometric mean (95% confidence interval) was 0.038 (0.024-0.060) _g/l (equivalent to 0.842 (0.530-1.34) pg/mg albumin). The highest value was 0.2 _g/l (4.43. pg/mg albumin). Based on liver function biomarkers, there was no evidence of increased liver dysfunction in these persons. Conclusions: During a time when exposure to aflatoxins in food products might have been expected to be increased, we identified few exposed persons. Although none of the subgroup analyses provided reliable estimates due to high relative standard errors, they suggested that additional targeted surveillance may be warranted. ¥1% had aflatoxin B1exposure.¥No evidence for liver dysfunction.¥Additional targeted surveillance may be warranted.