Preterm infants undergoing intensive care are often placed prone to improve respiratory function. Current clinical guidelines recommend preterm infants are slept supine from 32 weeks' postmenstrual age, regardless of gestational age at birth. However, respiratory function is also related to gestational and chronological ages and is affected by sleep state. We aimed to identify the optimal timing for adopting the supine sleeping position in preterm infants, using a longitudinal design assessing the effects of sleep position and state on cardiorespiratory stability.Twenty-three extremely (24-28 weeks' gestation) and 33 very preterm (29-34 weeks' gestation) infants were studied weekly from birth until discharge, in both prone and supine positions, in quiet and active sleep determined by behavioral scoring. Bradycardia (heart rate ≤100 bpm), desaturation (oxygen saturation ≤80%), and apnea (pause in respiratory rate ≥10 s) episodes were analyzed.Prone positioning in extremely preterm infants reduced the frequency of bradycardias and desaturations and duration of desaturations. In very preterm infants, prone positioning only reduced the frequency of desaturations. The position-related effects were not related to postmenstrual age. Quiet sleep in both preterm groups was associated with fewer bradycardias and desaturations, and also reduced durations of bradycardia and desaturations in the very preterm group.Cardiorespiratory stability is improved by the prone sleep position, predominantly in extremely preterm infants, and the improvements are not dependent on postmenstrual age. In very preterm infants, quiet sleep has a more marked effect than the prone position. This evidence should be considered in individualizing management of preterm infant positioning.