Objective: This study aimed to assess whether the use of levothyroxine was beneficial in pregnant women with subclinical hypothyroidism (SCH) or women who were positive for thyroid peroxidase antibody (TPOAb+). Study design: This was a randomized clinical trial involving 1736 populations of pregnant women. Our study was conducted into two groups: normal pregnant group and recurrent pregnancy loss (RPL) pregnant group. Both patients were divided into SCH and TPOAb+ groups, respectively; the patients of four groups were randomized into either the L-T4 treatment or control groups. Results: There was no significant difference between the two groups in the prevalence of SCH and TPOAb+ (p > .05); The live births rate of normal pregnant group was higher than that of the RPL pregnant Group (79.5% vs. 70.8%, p < .05); Considering normal pregnant women, whether women who were TPOAb + or SCH, there was no significant difference between the treatment and control group in the live birth rate and the pregnancy loss rate. In RPL group, whether women who were TPOAb + or SCH, L-T4 treatment can obtain higher live birth rate and lower abortion rate compared to the control. Conclusion: There was no significant difference between normal and RPL pregnant women in the prevalence of SCH and TPOAb+.Treatment with L-T4 decreased the risk of pregnancy loss and increased the live birth rate in RPL pregnant women who were positive for TPOAb or subclinical hypothyroidism. Levothyroxine therapy is recommended for SCH and TPOAb + women in pregnant women with recurrent pregnancy loss.