Objective To investigate the effect of physical activity (PA) on development (motor, cognitive, social-emotional) in children 4–5 years old born <30 weeks’ gestation, and to describe subgroups of children at risk of low PA in this cohort.

Design Longitudinal cohort study.

Patients 123 children born <30 weeks were recruited at birth and assessed between 4 and 5 years’ corrected age.

Main outcome measures Development was assessed using the Movement Assessment Battery for Children, Second Edition (MABC-2), Little Developmental Coordination Disorder Questionnaire (L-DCDQ), Wechsler Preschool and Primary Scale of Intelligence (Fourth Edition; WPPSI-IV), and Strengths and Difficulties Questionnaire (SDQ). To measure PA, children wore an accelerometer and parents completed a diary for 7 days. Effects of PA on developmental outcomes, and associations between perinatal risk factors and PA, were estimated using linear regression.

Results More accelerometer-measured PA was associated with better MABC-2 aiming and catching scores (average standard score increase per hour increase in PA: 0.54, 95% CI 0.11, 0.96; p=0.013), and lower WPPSI-IV processing speed index scores (average composite score decrease per hour increase in PA: −2.36, 95% CI −4.19 to –0.53; p=0.012). Higher accelerometer-measured PA was associated with better SDQ prosocial scores. Major brain injury in the neonatal period was associated with less moderate-vigorous and less unstructured PA at 4–5 years.

Conclusions Higher levels of PA are associated with aspects of motor, cognitive and social-emotional skill development in children 4–5 years old born <30 weeks. Those with major brain injury in the neonatal period may be more vulnerable to low PA at preschool age.