OBJECTIVE:: To estimate if there is a relationship between subclinical thyroid disease and gestational diabetes. METHODS:: Between November 2000 and April 2003, serum thyrotropin screening was performed on all women who presented for prenatal care. Women identified with abnormal thyrotropin had a serum free thyroxine reflexively determined. Those women with abnormal serum free thyroxine values were referred for further evaluation and excluded from further analysis. For this analysis, normal thyrotropin values were those between the 2.5th and 97.5th percentiles (0.03-4.13 milliunits/L) not corrected for gestational age and serum free thyroxine were considered normal if they ranged from 0.9 to 2.0 mg/dL. Women with an elevated serum thyrotropin but a normal serum free thyroxine were designated to have subclinical hypothyroidism and those with a low thyrotropin and a normal serum free thyroxine level were designated to have subclinical hyperthyroidism. Euthyroid women had both normal thyrotropin and normal serum free thyroxine values. The incidence of gestational diabetes was compared among these three groups. RESULTS:: Of the 24,883 women included in the study, 23,771 (95.5%) were euthyroid, 584 (2.3%) had subclinical hyperthyroidism, and 528 (2%) had subclinical hypothyroidism. The likelihood of gestational diabetes increased with thyrotropin level (P=.002). For example, when a pregnant Hispanic woman of average age and weight was used, the predicted percent of gestational diabetes increased from 1.9% to 4.9% as thyrotropin increased from 0.001 to 10 milliunits/L (P=.001). CONCLUSION:: The risk of developing gestational diabetes increases with thyrotropin level. This supports a relationship between subclinical hypothyroidism and diabetes diagnosed during pregnancy. Copyright © 2012 American College of Obstetricians and Gynecologists.