Background: The relation between folate status and cancer is controversial. Several epidemiologic studies have suggested that increased folate intake is associated with reduced risk of various cancers, others have found no such associations, and a few have suggested that high folate intake might increase the risk of certain cancers. Methods: Using data from the Third National Health and Nutrition Examination Survey (NHANES III) Mortality File, a prospective cohort study of a nationally representative sample of 14,611 U.S. adults, we conducted Cox proportional hazards regression modeling to investigate the association of baseline serum folate concentrations and all-cancer mortality determined from linked death certificate data. Results: Relative to the lowest quintile of serum folate (<3.0 ng/mL), the multivariable-adjusted hazard ratios across quintiles 2 to 5 were: 1.61 [95% confidence interval (95% CI), 1.11-2.32], 1.00 (95% CI, 0.65-1.49), 1.39 (95% CI, 0.96-2.03), and 0.85 (95% CI, 0.59-1.22). These findings did not differ substantially by age or sex, but the higher risk for those in the second quintile appeared limited to non-Hispanic whites. Conclusion: These findings suggest that there may be a nonlinear relationship between folate status and the risk of all-cancer mortality such that persons with low, but not grossly deficient, serum blood folate concentrations may be at increased risk. Further study is needed to determine whether these findings are due to chance, and if not, to clarify their biological basis.