Direct breastfeeding has numerous benefits and promoting successful breastfeeding is one of the most important local and global goals in neonatal care.1, 2 The World Health Organisation (WHO) goal is that 50% of the babies should be exclusively breastfed in the first 6 months.3 The Baby-Friendly Hospital Initiative (BFHI) aims to promote breastfeeding globally, and hospitals may apply for BFHI status.4 In 2020, the WHO published a preterm and neonatal care extension to the BFHI.5 This repeats Step 9 in the original version, which states that parents should be informed about the use and risks of pacifiers and bottles.6, 7 Although the wording has been softened since previous versions, and the benefits of all feeding strategies are stated, the advice is still not neutral.3 Furthermore, the BFHI material does not present the evidence well, in particular the quality and uncertainty of the evidence.4, 5, 8 However, the material does highlight the need to promote informed decision making by parents better, which is an improvement on previous versions. When we discuss the impact of pacifiers and bottles on breastfeeding, the most important outcome is direct breastfeeding, but many studies have also used breastmilk feeding. The BFHI has increased the breastmilk feeding rate9 in all income settings10, 11 and improved staff attitudes to breastfeeding.12 However, some studies have also reported downsides to the BFHI, such as increases in hypoglycaemia cases or prolonged hospital stays, although these findings could also have been due changed treatment guidelines.13, 14 It should also be noted that hygiene and clean water resources may be limited globally and may lead to increased risk of infections in neonates and infants if bottles and pacifiers cannot be cleaned adequately. Possible enteral feeding strategies in neonatal units include nasogastric tube feeding, cup feeding, finger feeding, bottle feeding and breastfeeding. A key criterion for discharge is successful feeding and parents being able to perform it at home. Nevertheless, evidence has indicated that preterm neonates benefit from non-nutritional sucking, such as with pacifiers, because it increases physiological stability and improves motor skills in the mouth.15 This different view discusses the evidence for BFHI Step 9. Bottle feeding has been criticised and controversial in neonatal care, as previous BFHI guidelines were against it and promoted alternative enteral feeding methods.8 Mothers have mixed feelings about bottle feeding and may feel guilt and anger if breastfeeding is unsuccessful. However, they have also reported lack of information and support from healthcare professionals about bottle feeding.16 The primary criticism is that bottle feeding is too efficient and easy for the neonate, making the sucking technique less optimal. It may cause more desaturation and lower temperatures in sick neonates.5 Nevertheless, practices, such as nasogastric tube feeding and cup feeding, do not promote sucking techniques or motor skills.5 Notably, mothers are encouraged to breastfeed before alternative strategies are used, which is possible with bottle feeding.5 Although researchers have compared direct breastfeeding rates between bottle-fed and non-bottle-fed neonates, the control groups have typically used cups. Only one study used nasogastric tube feeding as a control in a randomised study.17 This is disappointing, given that it is a common feeding practice and more invasive procedure. The only randomised trial that compared nasogastric tubes to bottles was in 1999, but the groups used by Kliethermes et al. were unbalanced, and the randomisation was not completely successful.18 For example, only 27% of the mothers in the bottle group had successfully breastfed compared with 56% in the nasogastric tube group. This introduced clear bias, as previous successful breastfeeding is a well-known strong predictor for later successful breastfeeding.19 Furthermore, a Cochrane review concluded that cup feeding improved breastfeeding rates at discharge and 6 months.17 However, the review did not define breastfeeding as direct breastfeeding or breastmilk feeding. A previous systematic review found that cup feeding was somewhat unpractical, due to spillage and decreased milk intake, which reduced willingness and acceptance.20 Bottle feeding has some advantages, including being convenient for parents to learn and manage. It may also reduce hospitalisation, although the evidence is controversial.21, 22 Bottle feeding sick neonates has been associated with desaturation and lower temperatures, which should be noted by neonatal units.5 The overall evidence on bottle feeding in neonatal care remains inadequate and clear evidence from a randomised study comparing nasogastric tube feeding and bottle feeding or exclusive direct breastfeeding is warranted before it is recommended. Pacifiers have been discouraged because many observational studies have reporting that their use or early introduction may weaken breastfeeding success rates.23 However, observational studies, especially retrospective ones, are prone to bias. For example, confounding, selection, reporting and/or recall bias have affected observational studies that have associated pacifier use to weaker breastfeeding success. Randomised trials have not linked pacifier use to lower breastfeeding success.24 Notably, Kramer et al. found that breastfeeding rates were similar between restricted and free pacifier use in an intention-to-treat analyses, but pacifier use was higher among those who did not breastfeed in a retrospective analysis.25 This indicated that restricting pacified use as an intervention was not effective but may have had efficacy. This is a typical problem in studies about interventions that aim to change behaviour or practice. Thus, pacifier use may not weaken breastfeeding rates, and there may be other challenging infant-related problems with breastfeeding. That study exemplifies the overall need for randomised controlled trials that analyse the effectiveness of interventions and why intention-to-treat is the gold standard. A meta-analysis published in 2022 reported that pacifier use was not associated with breastfeeding outcomes in preterm and term neonates,24 but it did not classify breastfeeding as direct breastfeeding. That was a clear weakness. The most recent Cochrane review, from 2016, reached a similar conclusion, although it had a similar weakness with regard to the breastfeeding definition.26 Pacifiers have been shown to improve oral function.27 The 2022 meta-analysis reported that introducing pacifiers to preterm born neonates shortened the duration from nasogastric tube to oral feeding and total hospitalisation time.24 Previous Cochrane reviews have also reported that non-nutritive sucking was beneficial for preterm neonates.15 Pacifiers also have other known benefits and harms not related to breastfeeding. The harms include increased acute otitis media and problems with teething, if continued for too long.28, 29 Pacifiers also have been shown to prevent sudden infant death syndrome in observational studies30 but so did breastfeeding.31 Thus, Step 9 regarding pacifiers in the WHO's BFHI seems rather harsh given the vague and low-quality underlying evidence from observational studies. Better quality evidence from randomised trials suggest that pacifiers do not harm breastfeeding. Thus, it would be reasonable to provide information on the benefits and harms to mothers. Evidence supporting the WHO's BFHI Step 9 remains vague. Better quality research is required to analyse the possible benefits and harms of bottle feeding before giving stronger recommendations for or against their use in neonatal care. Pacifiers are feared too much based on weak observational evidence, when randomised studies have not indicated they harm breastmilk feeding in term or preterm neonates. Although it must be noted that these studies focused on breastmilk feeding and not direct breastfeeding. My perspectives cannot be generalised directly to settings with limited resources, for example, a lack of clean water. Pacifiers and bottles can pose an infection risk in such settings, such as gastroenteritis, which remains one of the highest causes of global child mortality. My conclusion is that Step 9 should be more informative for healthcare professionals and caregivers. Mothers and parents should still be counselled about the possible benefits and harms of pacifiers and bottles when the evidence does not provide factual answers on direct breastfeeding. This would help to promote the shared decision making that is recommended in the BFHI material. None. I have no financial conflicts of interest. My first-born child did not have pacifiers and did not use a bottle. My second born child was introduced early to pacifiers and bottle-feeding already in the neonatal intensive care unit. Both have been breastfed as long their mother has wished to do so. At the moment, I work in neonatal unit without a baby-friendly hospital certificate. These factors might be seen as an intellectual conflict of interest.