Aim Subclinical hypothyroidism is thought to be associated with adverse pregnancy outcomes but the data is conflicting and generally depends on antibody positivity and treatment. We evaluated the pregnancy outcomes in Turkish population with untreated, antibody negative subclinical hypothyroidism for the first time. Methods We searched for 30 015 patients between January 2016 and May 2017 retrospectively. Finally, a total of 930 pregnant women with untreated, antibody negative subclinical hypothyroidism and 7986 controls were included. Demographic characteristics, laboratory findings and pregnancy outcomes, including pregnancy loss, impaired glucose tolerance, gestational diabetes, hypertensive disorders of pregnancy, preterm birth, neonatal intensive care unit admission, placenta previa and abruption, cesarean delivery, low birthweight, Apgar score Demographic and laboratory characteristics were similar between two groups except thyroid stimulating hormone levels and previous uterine surgery rates. Subclinical hypothyroidism group had an increased risk of pregnancy loss (odds ratio [OR] 2.583; 95% confidence interval [CI] 1.982-3.365; P < 0.001), impaired glucose tolerance (OR 1.952; 95% CI 1.450-2.627; P < 0.001), hypertensive disorders of pregnancy (OR 1.476; 95% CI 1.113-1.923; P = 0.004), neonatal intensive care unit admission (OR 1.620; 95% CI 1.084-2.420; P = 0.019), placenta previa (OR 12.581; 95% CI 5.046-31.363; P < 0.001) and cesarean delivery (OR 1.263; 95% CI 1.091-1.462; P = 0.002). Conclusion Subclinical hypothyroidism has worse pregnancy outcomes as compared to euthyroid pregnant women even in antibody negativity. Therefore, we suggest that all pregnant women should routinely be screened in their first antenatal visits for thyroid functions.