Background Prevention and treatment in preterm birth are still under intensive investigation. A gap exists between evidence-based recommendations and clinical practice. A deeper understanding of the prevalence of medication use is an essential step toward improving the care of pregnant women. This study aimed to address this issue.Methods A retrospective cohort was conducted between December 2018 and November 2019 in Tongji Hospital (Wuhan). In total, 878 pregnant women were enrolled. Information on maternal characteristics, medication prescription, maternal outcomes (including post-partum hemorrhage), and neonatal outcomes (including APGAR score, birthweight, neonatal intensive care unit (NICU) admission, cardiopulmonary and neurological diseases) in the linking system were retrieved.Results The overall coverage of the common medications, including antenatal corticosteroids (ACS), magnesium sulfate, tocolytics, and progesterone, was 84.5%, 52.8%, 24.9%, and 13.3%, respectively. The treatment plan varied widely regarding the regimen, dosing, duration, and combination of these drugs. Specifically, tocolytic maintenance, which means the administration duration of more than 48 hours in this study, increased the risk of post-partum hemorrhage; it also significantly increased birthweight, but was not a risk factor of large-for-gestational-age (LGA).Conclusions This retrospective study in a single tertiary center disclosed that obstetric medications for preterm labor are highly individualized and do not strictly conform to current guidelines. This phenomenon is not uncommon in other countries. Clinical practice and evidence-based recommendations need to be carefully balanced; more studies are warranted to provide high-quality evidence to aid clinical decisions.