To describe the exposure to drugs used for the treatment of patent ductus arteriosus (PDA) in a large cohort of preterm infants born before 32 weeks of gestation. METHODS: A prospective observational cohort analysis was conducted during 2 years in 28 French level 3 NICU using the same computerized order-entry system. The main outcome was "a medically treated PDA," defined as exposure to ibuprofen, indomethacin, or paracetamol prescribed with the indication of PDA closure. Secondary outcomes were as follows: time of the first treatment administration; total exposure to furosemide during hospitalization; and rate of PDA refractory to pharmacological closure. RESULTS: The study cohort consisted of 2614 infants. Among them, 474 (18.1%) received a medical treatment for PDA, with a mean postnatal age at treatment of 4.3 ± 6.6 days. The drug used as a first-line treatment was ibuprofen in 89.5% and paracetamol in 10.5%. One hundred and ninety-five infants (7.4%) had a PDA refractory to pharmacological closure. At the multivariate analysis, factors associated with PDA refractory to pharmacological closure (OR; 95% CI) were as follows: gestational age (GA) (0.81; 0.72-0.90), paracetamol as the first-line treatment (0.32; 0.15-0.68), and pharmacological treatment before 48 h of life (0.63; 0.43-0.94). 24.6% of the study cohort was exposed to furosemide (cumulative dose 6.5 ± 12.6 mg/kg). Variables significantly associated with higher cumulative doses of furosemide were lower GA and ibuprofen treatment (both p < 0.0001). CONCLUSION: Drug utilization patterns in infants with PDA vary among centers. Pharmacoepidemiology studies can provide new information on factors associated with PDA refractory to medical treatment.