ObjectiveTo (1) describe differences in types and timing of interventions, (2) report short-term outcomes and (3) describe differences among centres from a large national cohort of preterm infants with post-haemorrhagic hydrocephalus (PHH).DesignCohort study of the Children’s Hospitals Neonatal Database from 2010 to 2022.Setting41 referral neonatal intensive care units (NICUs) in North America.PatientsInfants born before 32 weeks’ gestation with PHH defined as acquired hydrocephalus with intraventricular haemorrhage.Interventions(1) No intervention, (2) temporising device (TD) only, (3) initial permanent shunt (PS) and (4) TD followed by PS (TD-PS).Main outcome measuresMortality and meningitis.ResultsOf 3883 infants with PHH from 41 centres, 36% had no surgical intervention, 16% had a TD only, 19% had a PS only and 30% had a TD-PS. Of the 46% of infants with TDs, 76% were reservoirs; 66% of infants with TDs required PS placement. The percent of infants with PHH receiving ventricular access device placement differed by centre, ranging from 4% to 79% (p<0.001). Median chronological and postmenstrual age at time of TD placement were similar between infants with only TD and those with TD-PS. Infants with TD-PS were older and larger than those with only PS at time of PS placement. Death before NICU discharge occurred in 12% of infants, usually due to redirection of care. Meningitis occurred in 11% of the cohort.ConclusionsThere was significant intercentre variation in rate of intervention, which may reflect variability in care or referral patterns. Rate of PS placement in infants with TDs was 66%.