Background: Tocopherols were discovered for their role in animal reproduction, but little is known about the contribution of deficiencies of vitamin E to human pregnancy loss. Objective: We sought to determine whether higher first-trimester concentrations of α-tocopherol and γ-tocopherol were associated with reduced odds of miscarriage (pregnancy losses <24 wk of gestation) in women in rural Bangladesh. Design: A case-cohort study in 1605 pregnant Bangladeshi women median (IQR) gestational age: 10 wk (8–13 wk) who participated in a placebo-controlled vitamin A– or β-carotene–supplementation trial was done to assess ORs of miscarriage in women with low α-tocopherol (<12.0 μmol/L) and γ-tocopherol (<0.81 μmol/L; upper tertile cutoff of the γ-tocopherol distribution in women who did not miscarry). Results: In all women, plasma α- and γ-tocopherol concentrations were low median (IQR): 10.04 μmol/L (8.07–12.35 μmol/L) and 0.66 μmol/L (0.50–0.95 μmol/L), respectively. In a logistic regression analysis that was adjusted for cholesterol and the other tocopherol, low α-tocopherol was associated with an OR of 1.83 (95% CI: 1.04, 3.20), whereas a low γ-tocopherol concentration was associated with an OR of 0.62 (95% CI: 0.41, 0.93) for miscarriage. Subgroup analyses revealed that opposing ORs were evident only in women with BMI (in kg/m ²) ≥18.5 and serum ferritin concentration ≤150 μg/L, although low BMI and elevated ferritin conferred stronger risk of miscarriage. Conclusions: In pregnant women in rural Bangladesh, low plasma α-tocopherol was associated with increased risk of miscarriage, and low γ-tocopherol was associated with decreased risk of miscarriage. Maternal vitamin E status in the first trimester may influence risk of early pregnancy loss. The JiVitA-1 study, from which data for this report were derived, was registered at clinicaltrials.gov as NCT00198822.