ObjectiveWhen medically indicated, caesarean section can prevent deaths and other serious complications in mothers and babies. Lack of access to caesarean section may result in increased maternal and perinatal mortality and morbidity. However, rising caesarean section rates globally suggest overuse in healthy women and babies, with consequent iatrogenic damage for women and babies, and adverse impacts on the sustainability of maternity care provision. To date, interventions to ensure that caesarean section is appropriately used have not reversed the upward trend in rates. Qualitative evidence has the potential to explain why and how interventions may or may not work in specific contexts. We aimed to establish stakeholders’ views on the barriers and facilitators to non-clinical interventions targeted at organizations, facilities and systems, to reduce unnecessary caesarean section.MethodsWe undertook a systematic qualitative evidence synthesis using a five-stage modified, meta-ethnography approach. We searched MEDLINE, CINAHL, PsychINFO, EMBASE and grey literature databases (Global Index Medicus, POPLINE, AJOL) using pre-defined terms. Inclusion criteria were qualitative and mixed-method studies, investigating any non-clinical intervention to reduce caesarean section, in any setting and language, published after 1984. Study quality was assessed prior to data extraction. Interpretive thematic synthesis was undertaken using a barriers and facilitators lens. Confidence in the resulting Summaries of Findings was assessed using GRADE-CERQual.Results8,219 studies were identified. 25 studies were included, from 17 countries, published between 1993–2016, encompassing the views of over 1,565 stakeholders. Nineteen Summary of Findings statements were derived. They mapped onto three distinct themes:Health system, organizational and structural factors (6 SoFs); Human and cultural factors (7 SoFs); and Mechanisms of effect to achieve change factors (6 SoFs). The synthesis showed how inter- and intra-system power differentials, and stakeholder commitment, exert strong mechanisms of effect on caesarean section rates, independent of the theoretical efficacy of specific interventions to reduce them.ConclusionsNon-clinical interventions to reduce caesarean section are strongly mediated by organisational power differentials and stakeholder commitment. Barriers may be greatest where implementation plans contradict system and cultural norms.Protocol registrationPROSPERO: CRD42017059456