PURPOSE:Pancreatic cancer(PCa) is one of the most lethal cancers with few known consistent nutrition-related risk factors. Epidemiologic associations between the trace element selenium and PCa are inconsistent. This study examined the association of pre-diagnostic serum selenium with incident PCa. METHODS:We conducted a nested case-control study within the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Study (PLCO) cohort of men and women 55-70 years old at baseline (1993-2001). In total, 303 PCa cases developed during the 17-year follow-up period (1993-2009). We selected two controls (n = 606) for each case who were alive at the time the case was diagnosed who were matched on age, sex, race, and date of blood draw. We used conditional logistic regression analysis to calculate the odds ratio (OR) and 95% confidence intervals (CI) adjusting for smoking status and diabetes mellitus. RESULTS:Mean serum selenium concentrations were slightly lower in cases (mean, 95% CI: 139.0 ng/ml, 135.6-138.9) compared to controls (142.5 ng/ml, 140.4-142.4, p = 0.08). Overall, serum selenium was not associated with PCa risk (continuous OR: 0.66; 0.32-1.37). There was no significant interaction by sex, smoking, diabetes, or follow-up time (p > 0.05). CONCLUSION:Our results do not support the hypothesis that serum selenium is associated with PCa risk.