Objective: To investigate if women with subclinical hypothyroidism (SCH), positive thyroid gland peroxidase antibody (TPOAb) in early pregnancy accepted treatment or not had effect on perinatal outcomes., Methods: 15 000 pregnant women who delivered in Women and Infants Hospital of Zhengzhou from January 1, 2013 to June 30, 2014 were recruited retrospectively. Among them, 2 042 women had SCH in early pregnancy. The diagnostic standard of SCH was serum free thyroxine (FT4) between 12.91-22.35 pmol/L and TSH level between 5.22-10.00 mU/L. TPOAb level ≥ 34 U/L was defined as positive result. The 2 042 patients with SCH were divided into the treated group (1 236 cases) and the untreated group (806 cases), according to whether or not women accepted the levothyroxine treatment. Meanwhile, the 2 042 patients with SCH were divided into the TPOAb (+) treated group (1 021 cases), the TPOAb (+) untreated group (201 cases), the TPOAb (-) treated group (215 cases) and the TPOAb (-) untreated group (605 cases), according to the TPOAb result and acceptance the levothyroxine treatment. 2 000 pregnant women with normal thyroid function who delivered in the same period were selected as the control group. Perinatal outcomes were analyzed., Results: (1) The incidence of SCH in early pregnancy was 13.61% (2 042/15 000). 60.53% (1 236/2 042) accepted levothyroxine treatment and 39.47% (806/2 042) did not. (2) The incidence of abortion (5.71%, 46/806), premature delivery (6.20%, 50/806), gestational hypertension disease (13.90%, 112/806), gestational diabetes mellitus (GDM; 6.58%, 53/806), fetal growth restriction (FGR; 12.28%, 99/806) and low birth weight infants (10.17%, 82/806) in the untreated group were higher than those in the treated group [3.96% (49/1 236), 4.21% (52/1 236), 10.76% (133/1 236), 4.13% (51/1 236), 8.90% (110/1 236), 7.52% (93/1 236), respectively] and the control group [3.60% (72/2 000), 4.00% (80/2 000), 10.70% (214/2 000), 3.80% (76/2 000), 9.60% (192/2 000), 7.50% (150/2 000), respectively]. The differences were statistically significant (P < 0.05). While there was no statistically significant difference in the incidence of placental abruption, anemia in pregnant women, or fetal distress among the three groups (P > 0.05). (3)The incidences of abortion (11.44%, 23/201), premature delivery (12.44%, 25/201), gestational hypertension disease (22.89%, 46/201), GDM (8.46%, 17/201), FGR (19.90%, 40/201) and low birth weight infants (16.42%, 33/201) in the TPOAb (+) untreated group were higher than those in TPOAb (+) treated group [4.02% (41/1 021), 4.21% (43/1 021), 10.77% (110/1 021), 4.11% (42/1 021), 8.72% (89/1 021), 7.35% (75/1 021), respectively] and the control group, with statistically significant differences (P < 0.05). The incidence of the pregnancy complications in the TPOAb (+) treated group was higher than those in the control group, but the differences were not statistically significant (P > 0.05). (4) There were no statistically significant difference (P > 0.05) in the incidence of abortion (3.72%, 8/215), premature delivery (4.19%, 9/215), gestational hypertension disease (10.70%, 23/215), GDM (4.19%, 9/215), FGR (9.77%, 21/215) or low birth weight infants (8.37%, 18/215) among the TPOAb (-) treated group, the TPOAb (-) untreated group [3.80% (23/605), 4.13% (25/605), 10.91% (66/605), 5.95% (36/605), 9.75% (59/605), 8.10% (49/605), respectively] and the control group., Conclusions: (1) The incidence of abortion, premature delivery, gestational hypertension disease, GDM, FGR and low birth weight infants could be increased in women with SCH in early pregnancy. (2) Thyroxine treatment could reduce the incidence of pregnancy complications in women with SCH in early pregnancy.