Files included Documentation_16102022 : Excel file describing the data with all variables labelled and units of measurement and formats provided Treatment_BP_Comparison_16102022 : Data file in CSV format Study design: This was a retrospective medical records review study. The neonatal unit outcomes of all extreme preterm babies were compared between Hospital 1 which aims to maintain BP above 30 mmHg irrespective of the gestational age (GA) (Active BP support group) and Hospital 2 which uses medication only if babies developed clinical signs of low BP (Permissive BP support group). The criteria used for intervention in the Permissive BP support group was mean arterial blood pressure (MABP) lower than the gestational age in weeks of the infant with clinical evidence of poor perfusion (poor skin colour/capillary refill time >3 secs, urine output <1ml/kg/hour from weighing of urine, lactate >3mmol/l, worsening base deficit/base deficit >8mmol/l and/or increasing oxygen requirement). Both centres had a comparable BP management protocol, consisting of initial 0.9% saline bolus of 10 to 20 ml/kg, followed by Dopamine, and then the addition of other inotropes if required, to achieve the predetermined Active and Permissive BP support. Participants: All preterm infants born at 23+0-28+6 weeks gestation admitted to two neonatal units over 4 years from 1st January 2007 to 31st December 2010 were eligible. Babies admitted after 12 hours of age, with major congenital abnormalities, and whose intensive care observational charts (to collect BP data) were not traceable, were excluded. Data collected: Pregnancy complications (antenatal Doppler findings, chorioamnionitis, hypertension), condition of the baby at birth (Clinical Risk Index for Babies (CRIB) score, cord pH and Lactate), and demographics of infants were collected using Electronic Patient Record (EPR) and BadgerNet (UK national neonatal EPR). Hourly systolic (SBP) and mean arterial (MABP) BP data were collected from intensive care charts. Inotropic medications, fluid bolus, red blood cell and other blood products used were gathered from medication and/or prescription charts. Short-term outcomes such as necrotising enterocolitis (NEC), intraventricular haemorrhage (IVH)/brain injury identified by brain ultrasound scan, bronchopulmonary dysplasia (BPD), patent ductus arteriosus (PDA), death or discharge, and duration of BAPM levels of care provided were collected from BadgerNet. NEC was defined by Bell’s stage of classification 23. BPD was defined as oxygen dependency at 36 weeks post conceptional age 24. PDA was confirmed by echocardiography. Ten percent of BP data collected was randomly verified by two researchers for confirmation of accuracy.