Objective: We attempted to evaluate maternal thyroid function in a new self-sequential longitudinal reference interval (SLRI) which we established recently. By this method, we analysed the correlation between pregnancy outcome, neonatal thyroid stimulating hormone (TSH) level and maternal thyroid diseases. Methods: A total of 1744 pregnant women participated in the study and 1747 babies were born from those women (three bore twins). The levels of TSH, free thyroxine (FT4) and thyroid peroxidase antibodies (TPO-Ab) of mothers were quantified by electrochemistry immunoassay (ECL). The levels of neonatal blood TSH were detected by time-resolved fluorescence immunoassay (TRFIA). All data were collected and statistically analysed by SPSS 13.0 software. Results: With our new SLRI method, we found that 0.11%~3.84% pregnant women would get thyroid diseases. Subclinical hypothyroidism was the most common maternal thyroid disorder. Being positive for thyroid peroxidase antibodies was a significant risk factor of subclinical hypothyroidism during pregnancy. The median, P2.5~P97.5, and interquartile range (IQR) of neonatal TSH (N-TSH) of 1747 babies were 2.72 mIU/L, 0.10~8.01 mIU/L and 2.62 mIU/L, respectively; 28.6% of pregnant women with thyroid diseases developed pregnancy complications. The prevalence was significantly higher than in the normal thyroid function group (p < 0.001). The levels of N-TSH were low correlated with maternal TSH levels (p < 0.05), but there were no significant correlations between N-TSH and maternal FT4 and maternal TPO-Ab (p > 0.05). Conclusions: Thyroid disorders, especially subclinical hypothyroidism, are common in pregnant women. These disorders are associated with pregnancy and fetal outcome. Routine maternal thyroid function screening is important and should be recommended.