Sleep is paramount for optimal brain development in infants admitted to the neonatal intensive care unit. Besides (minimally) invasive technical approaches to study sleep in infants, there is currently a large variety of behavioral sleep stage classification methods (BSSCs) that can be used to identify sleep stages in preterm infants born <37 weeks gestational age. However, they operate different criteria to define sleep stages, which limits the comparability and reproducibility of research on preterm sleep. This scoping review aims to: 1) identify and elaborate on existing neonatal BSSCs used for preterm infants, 2) examine the reliability and validity of these BSSCs, and 3) identify which criteria are most used for different ages, ranging from 23 to 37 weeks postmenstrual age at observation. To map the existing BSSCs, PubMed, EMBASE and Cochrane were searched for studies using a BSSC to identify sleep stages in preterm infants. In total, 36 BSSCs were identified with on average five item categories assessed per BSSC, most frequently: eyes, body movements, facial movements, sounds, and respiratory pattern. Furthermore, validity and reliability of the BSSCs were tested in less than half of the included studies. Finally, BSSCs were used in infants of all ages, regardless the age for which the BSSC was originally developed. Items used for scoring in the different BSSCs were relatively consistent. The age ranges, reliability, and validity of the BSSCs were not consistently reported in most studies. Either validation studies of existing BSSCs or new BSSCs are necessary to improve the comparability and reproducibility of previous and future preterm behavioral sleep studies. •Preterm sleep stages can be assessed using behavioral classification methods (BSSCs).•Many BSSCs used in preterm infants are originally developed for term infants.•Validity and reliability are not consistently reported in BSSCs used in preterm infants.•Validated BSSCs are necessary to improve reproducibility of preterm sleep studies.•Most BSSCs comprised five frequently recurring categories of scoring items.