Preterm birth increases the risk for neurologic and developmental disabilities and therefore long-term follow-up is important. This prospective follow-up study aims to describe longitudinal motor performance in preterm infants from 6 to 24 months and to detect the influence of risk factors on motor performance trajectories. We included preterm infants (n = 348) with a gestational age of [less than or equal] 32 weeks. The Bayley Scales of Infant Development, 2nd edition (BSID-II) Motor Scale and the Behaviour Rating Scale were recorded at the corrected ages of 6, 12 and 24 months. The Motor Scale raw score was the dependent variable in random coefficient analysis for risk factors in the cohort if infants with cerebral damage were in- and excluded. The raw score increased, showed the highest correlation (r[subscript p] = 0.48-0.67) and was more stable than the PDI and its classification. Fifteen percent of the infants had a stable classification, while 45% changed one class. Male sex and intra-ventricular haemorrhage (IVH) lowered the raw scores. Higher motor quality scores and height increased the raw scores, while the influence of maternal education varied at different time points. Removal of infants with cerebral damage from the cohort did not change the risk factors. The results showed that the raw score trajectories were more stable, but after corrections for norm data, the measurements became highly unstable. This is clinically important when reporting results to parents, guiding intervention and in randomised trials. The risk factors predominantly influenced the level of motor performance raw scores. Maternal education additionally influenced the trajectory and should be included in randomisation procedures. (Contains 3 figures and 2 tables.)