Despite some studies indicating that thyroid antibody positivity during pregnancy has been associated with adverse pregnancy outcomes, evidence regarding the effects of levothyroxine (LT4) treatment of euthyroid/subclinical hypothyroid pregnant women with autoimmune thyroid disease on pregnancy outcome is limited. We aimed to assess whether pregnant women with autoimmune thyroid disease, but without overt thyroid dysfunction are affected by higher rates of adverse pregnancy outcomes. In addition, we aimed to explore whether LT4 treatment improves the pregnancy outcome of affected women. A prospective study was carried out on pregnant women from the first trimester to delivery. The study was conducted among pregnant women receiving prenatal care in centers under coverage of Shahid Beheshti University of Medical Sciences. Of a total of 1746 pregnant women, screened for thyroid dysfunction, 1028 euthyroid TPOAb-negative (TPOAb-) and 131 thyroid peroxidase antibody-positive (TPOAb+) women without overt thyroid dysfunction entered the second phase of the study. TPOAb+ women were randomly divided into two groups: group A (n = 65), treated with LT4 and group B (n = 66), received no treatment. The 1028 TPOAb- women (group C) served as a normal population control group. Primary outcomes were preterm delivery and miscarriage and secondary outcomes included placenta abruption, still birth, neonatal admission and neonatal TSH levels. Groups A and C displayed a lower rate of preterm deliveries compared with group B (RR = 0.30, 95% CI: 0.1-0.85, P = 0.0229) and (RR = 0.23, 95% CI: 0.14-0.40, P < 0.001) respectively. There was no statistically significant difference in the rates of preterm labor between groups A and C (RR = 0.79, 95% CI: 0.30-2.09, P = 0.64). The number needed to treat (NNT) for preterm birth was 5.9 (95% CI: 3.33–25.16). Treatment with LT4 decreases the risk of preterm delivery in women who are positive for TPOAb.