BACKGROUND: There have been no recent US population-based estimates of syphilis seroprevalence. We determined the prevalence of syphilis seroreactivity among a representative sample of the US population. METHODS: Sera from 18- to 49-year-old participants in the National Health and Nutrition Examination Surveys 2001-2004 were tested for syphilis IgG antibody using an enzyme immunoassay (EIA). Specimens with positive or indeterminate EIAs underwent rapid plasma reagin (RPR) testing; RPR titers > or =1:8 were considered positive. Specimens with RPR titers <1:8 underwent confirmatory testing with Treponema pallidum particle agglutination (TP-PA). RESULTS: Sera were available for 5767 participants. EIA testing was positive or indeterminate for 126, of which 10 had RPR titers > or =1:8. Of the remaining 116 specimens, 60 had positive TP-PA tests, including all 19 with RPR titers >1:1. Overall weighted syphilis seroprevalence was 0.71% (95% CI: 0.51-0.96). Prevalence was similar among males (0.76%) and females (0.67%) and increased with age, less education, and lower income (P <0.001 for each). Non-Hispanic blacks had the highest prevalence (4.3%), followed by Mexican-Americans (0.98%) and non-Hispanic whites (0.07%; P <0.001). CONCLUSIONS: The prevalence of syphilis seroreactivity was low (0.71%) in the general US population of 18- to 49-year-olds. However, consistent with surveillance data, this nationally representative survey showed substantial disparities in syphilis by race/ethnicity.