Subclinical hypothyroidism (SCH) is defined by elevated thyrotropin (TSH) and normal level of thyroxine (T4). The definition of SCH and the cutoff for TSH normality in pregnancy are debated. In the present study, we assess offspring perinatal outcome, anthropometrics and early development in relation to different TSH levels. An observational study with 77 singleton-pregnant women included by thyroid screening before a planned cesarean section. Two TSH-cutoffs (3.0 and 3.7 mIU/L) defined euthyroid and SCH groups, and were applied to evaluate offspring anthropometrics, complication rates (maternal blood loss, Apgar-score, cord arterial-pH, admission to neonatal intensive care unit, perinatal hypoglycemia) and offspring development. Development was evaluated by Bayley-III test in a subsample at age 6 months (n=27) and 15 months (n=22). Prevalence of SCH was 31.2 % at TSH-cutoff 3.0 mIU/L, and 16.9 % at TSH-cutoff 3.7 mIU/L. No differences in complications and anthropometrics were observed. In Bayley-III tests, cognitive score was decreased at 6 months (p=0.012) and at 15 months (p=0.056) by applying TSH-cutoff 3.0 mIU/L. At cutoff 3.7 mIU/L, motor score was decreased at 15 months (p=0.020). Male offspring had significantly lower cognitive scores at age 6 and 15 months (TSH-cutoff 3.0 mIU/L), and motor scores at age 15 months (TSH-cutoff 3.7 mIU/L). The importance of the definition of thyroid normality in pregnancy is underlined. This study suggests that a gender-effect might be present in maternal thyroid disease, and that developmental differences exist if TSH-cutoff is low. Further research is needed.