Objective To evaluate whether a perfusion-based approach (permissive hypotension, PH) for idiopathic low mean arterial blood pressure (MABP) during the first 72 hours after birth in preterm infants without overt sepsis affects neurodevelopmental outcome (NDO) at 24 months of age. Design Randomised controlled trial. Outcome assessors were blinded. Setting Single centre. Patients Infants <30 weeks gestational age (GA) with MABP in mm Hg <GA in weeks during the first 72 hours after birth, without overt signs of sepsis. Intervention Random assignment to PH, initiating treatment on signs of low perfusion and/or a MABP 5 mm Hg below their GA in weeks, or standard treatment (ST), initiating treatment when MABP was <GA in weeks. Main outcome measures Death, NDO at 24 months, and composite adverse outcome (death or cognitive and/or motor NDO below –1 SD). Results 86 infants were included, 57.3% of projected inclusions. Both cognitive NDO (PH-ST mean difference 0.8 (95% CI −5.6 to 7.3)) and motor NDO (mean difference 3.7 (−3.3 to 10.7)) were comparable. The relative risks for death (1.4 (0.6 to 3.7)) and composite adverse outcome (0.8 (0.5 to 1.3)) were comparable. The number of infants with inotropic support (n=19 (42.5%) vs 7 (15.2%), p=0.004) and duration of support (median 48.0 hours (IQR 26.8 to 77.5) vs 17.0 (14.0 to 37.0)) was lower in the PH group, with comparable MABPs. Conclusion A PH approach in preterm infants is feasible. It leads to comparable blood pressures with less inotrope administration. Though underpowered, we did not detect a major negative impact of PH on short-term or long-term outcomes. Trial registration number NCT01434251