INTRODUCTION: Addressing non-communicable diseases (NCDs) in humanitarian crises is increasingly relevant as the global NCD burden grows in tandem with rising rates of forced displacement. When the Syrian crisis began in 2011, there were gaps in research, tools, and guidelines to support NCD care delivery in crises. This thesis summarises four publications evaluating a Médecins sans Frontières (MSF) NCD programme for Syrian refugee and host populations in Irbid, Jordan. The evaluation sought to support service improvement, knowledge generation and translation to other settings. The objectives of this analytic commentary are to: 1. Summarise and critically appraise the methods and key findings of the evaluation and four related publications. 2. Situate the key findings within the broader policy and operational context and within the relevant literature. 3. Discuss implications of the findings for research on NCDs in humanitarian settings and for my own work. METHODS: A mixed methods study, guided by the RE-AIM implementation framework, comprised routine cohort data analysis, qualitative and descriptive costing studies, clinical audit, medication adherence survey and secondary analysis of an MSF household survey. KEY FINDINGS: This complex, multidisciplinary, vertical MSF NCD programme reached 25% of the target population. It was considered acceptable and accessible to patients, staff, and stakeholders; clinical guidelines were usable. Effectiveness was demonstrated by good clinical control and low defaulter rates. Implementation challenges included Syrian refugees’ social suffering and sense of hopelessness, which limited their capacity to engage in self-care, and proved challenging for staff. Programme adaptations included introducing mental health and psychosocial support (MHPSS), humanitarian liaison and home visit services. However, MHPSS service uptake was limited by low patient awareness, doctors’ distrust of the service, and mental health stigma. The programme’s costs, driven by human resources and medications, rose as its complexity increased. Programme maintenance was hindered by cost, short-term humanitarian funding and planning cycles, and limited integration within host systems. CONCLUSIONS: NCD programmes in humanitarian crises should be context-adapted and patient-centred, providing a continuum of care, including MHPSS and referral services, and catering for the range of clinical complexity within an NCD patient cohort. They should strive to be cost efficient while affordable and be designed in a participatory way, taking health system strengthening and sustainability perspectives to maintain continuity of care for the patient cohort, where possible. This evaluation built on previous cohort analyses, presented the first micro-costing study and MARS-5 medication adherence study, and added to the sparse published qualitative literature in this area. It contributed methodologically to indicator development, and conceptually to a framework for NCD care models in humanitarian settings. Its findings may support other actors engaged in NCD care in humanitarian crises.