Polysomnography is the gold standard for studying sleep, but it is complex to use, and this can be problematic in clinically unstable preterm infants. We evaluated the reliability of actigraphy and polysomnography in detecting sleep-wake patterns in newborn infants. A prospective, monocentric study was conducted that measured the sleep patterns of 48 infants: 24 late preterm neonates born at 34-36 weeks of gestational age and 24 term neonates. We used both polysomnography and the Actiwatch Mini during a three-hour period and then compared the results from the two methods. The baseline measurements for the preterm and terms groups were as follows: gestational age (34.5 weeks and 39.2 weeks), birthweight (2368 g and 3393 g) and age (6.4 days and 0.72 days). With the Actiwatch Mini, sensitivity for the late preterm and full-term infants was 78% and 87% for the leg actigraph and 78% and 93% for the arm actigraph. For specificity, the respective figures were 42% and 31% for the leg and 34% and 20% for the arm. Actigraphy using the Actiwatch Mini was not a reliable method for measuring sleep patterns in healthy late preterm and term neonates a few days after birth.