Philosophiae Doctor - PhD / The value of accountability as a key feature of strengthening health systems and reducing maternal, newborn and child mortality is increasingly emphasised globally, nationally and locally. Frontline health professionals and managers play a crucial role in promoting maternal, newborn and child health (MNCH) services in an equitable and accountable manner. They are at the interface between higher-level health system management and communities, facing demands from both sides and often expected to perform beyond their available means. Although accountability is a central topic in the governance of MNCH literature, it has mostly been approached at global and national levels, with little understanding of how accountability is integrated into the routine functioning of local health systems. This PhD explores the forms and functioning of accountability at the district level focusing on MNCH as a programmatic area with long-established institutional mechanisms (structures and processes) in South Africa (SA). The thesis is presented in the form of four empirical papers (published or submitted), exploring different dimensions of accountability, which are embedded in a series of narrative chapters. In this thesis, accountability is understood as a set of relations between an accountholder and ‘accountor’ (or duty bearer), in which the latter provides information or justification for actions or decisions taken, and faces the resulting consequences of his/her actions (reward or sanction). Accountability mechanisms are the means to regulate accountability relationships and include broad strategies, interventions or instruments. These mechanisms can take various forms including performance, financial and public accountability, and operate both vertically (accountability inside bureaucratic hierarchies, or towards external stakeholders and/or the community), or horizontally (between peers, ‘neighbour’ units, departments or ministries in a national health system). Drawing conceptually on the field of governance and considering the complexity of the accountability phenomenon, I adopted a case study approach to the PhD research, using a combination of policy document review, interviews (with managers, providers, community representatives and members of labour unions) and field observations, conducted iteratively over 16 months. The study was conducted in Gert Sibande District, one of the three South African health districts in Mpumalanga Province, with an in-depth focus on two of the seven sub-districts in the District. The research found that frontline health professionals have a clear understanding and conceptualisation of accountability in the SA health policy context, despite the reported inability to define accountability by health professionals described in the literature. Respondents referred to accountability as responsibility, answerability and virtue, and also argued for strengthening accountability mechanisms as critical to addressing maternal and child mortality. While deeming accountability as important, frontline professionals experienced the existing accountability mechanisms as ‘too much’ and indicated the desire for the streamlining of existing mechanisms. In this regard, the study documented numerous mechanisms at district level, almost all related to performance accountability in MNCH. These included a performance management system, quality assessment and accreditation processes, quarterly reviews, and death surveillance and response processes. The existence of multiple and overlapping accountability mechanisms engenders operational confusion and ‘accountability overload’ for frontline providers, encouraging empty bureaucratic compliance, while critical gaps – notably in community accountability – remain. In practice, at their best, some mechanisms operate following a reciprocal1 pathway of capacity building with resource provision (from management) and expectation for better performance (from providers). There were, however, contextual variations in the implementation and practice of the mechanisms between sub-district settings. The fieldwork observations and interviews were also able to document how formal institutionalised mechanisms are embedded within a complex system of informal accountability relationships and social norms (‘accountability ecosystem’) that enables or constrains the ability of frontline professionals to fulfil their tasks. In addition, using a Social Network Analysis approach, the research identified key actors and their involved network, which form the relational backdrop to the functioning of accountability mechanisms for MNCH. By revealing complex relationships and collaboration patterns among frontline health professionals, the study was able to show the multi-level action and multiple actors required to achieve MNCH goals.