Background: Approximately 1 million newborn deaths could be prevented each year through high quality health systems. Quality improvement collaboratives, involving group problem-solving in health facilities and sharing of learning across teams, are a widely used strategy to improve quality. However, rigorous evaluations are scarce; contextual influences and mechanisms of change are poorly understood, and the feasibility of using this strategy at scale is underexplored. Methods: I conducted a systematic review to understand how and under what circumstances quality improvement collaboratives may improve outcomes. I also evaluated the Safe Care Saving Lives programme, a quality improvement collaborative to reduce stillbirths and newborn mortality in 60 hospitals in Telangana and Andhra Pradesh, India. Using mixed methods, I evaluated impact on stillbirths and newborn mortality; contextual influences and mechanisms of change; and the feasibility of scaling up quality improvement through the state-level health insurance scheme which participating hospitals were part of. Results: Quality improvement collaboratives may affect outcomes through the normalisation of new behaviours and ways of working among clinical teams, supported by leaders. The evaluation of Safe Care Saving Lives found no effect on stillbirths and newborn mortality; high attrition in programme implementation due to the challenge of engaging leaders in quality improvement, and diluted implementation of the quality improvement collaborative approach. Scaling up the collaborative quality improvement approach through the state health insurance scheme in Telangana was not feasible. Barriers included limited coherence between the approach and the quality policy framework, and the limited scope of leveraging health insurance payments as incentives for quality in newborn care. Conclusion: this PhD highlights the limitations of quality improvement at facility level. Design of quality improvement should better consider strategies to engage leaders and respond to the needs of clinical teams, using formative research and theory of change. Greater coherence between quality improvement and other reform for quality in the health systems can aid improvement of newborn care at scale.