Complications of prematurity are the largest cause of childhood deaths globally, with >1 million deaths and high risk of long-term neurodevelopmental impairment. The first day after birth is the period of greatest risk and greatest potential for improving outcomes, especially with hospital-based small and sick newborn care (SSNC). Kangaroo mother care (KMC) involves continuous skin-to-skin contact and is recommended by WHO for all stable neonates ≤2000g. At PhD onset a priority evidence gap existed for initiation of early KMC before stability, which may contribute towards reducing the Sustainable Development Goal (SDG) for neonatal mortality, especially in low-resource, high-mortality contexts. My PhD aimed to investigate early KMC (< 24h of admission) in unstable neonates <2000g in one Gambian level 2/2+ neonatal unit. There were four objectives: (1) To prepare the research site with mitigation of barriers to trial implementation; (2) To investigate the effect of early KMC on survival and other clinical outcomes and safety (eKMC trial); (3) To explore pathways to preterm mortality and the effect of early KMC on physiological factors; (4) To determine the programmatic/policy and research implications of the PhD findings. The first PhD section provides the rationale for studying this topic and description of the local study site context at PhD onset. Substantial site preparations were required for trial implementation, including establishment of KMC as standard care and development/implementation of SSNC guidelines to minimise bias. Local data informed trial protocol development along with a conceptual framework to guide implementation process data collection. Unavailability of the mother during the first 24h of admission was a key recruitment barrier, mitigated by involving female relatives, and informed by a qualitative study to understand their perceptions towards SSNC and KMC. The second section focuses on the eKMC randomised controlled trial primary and secondary outcome findings. Despite adequate power at trial onset, no evidence of 28-day mortality effect was identified, with possible reasons including (1) Insufficient sample size to detect a between-arm difference due to large reductions in control arm mortality compared to pre-trial mortality (2) Low fidelity of the intervention delivered. However, important insights were gained for secondary outcomes, and feasibility of delivering prolonged KMC contact to unstable neonates, along with safety considerations. The final section presents a conceptual framework to describe pathways to mortality for neonates <2000g and potential amelioration by early KMC. Exploratory analyses of eKMC trial data identified substantial survival gains during the trial period, with 24% relative mortality reduction for all neonates <2000g and 29% relative risk reduction associated with trial participation. Weight <1200g, factors. This PhD provides valuable insights into SSNC in a West African context, underlining the importance of improving quality of SSNC overall as well as the potential for KMC as an entry point for family centred care. Female relatives are key stakeholders for family integrated SSNC and KMC in this socio-cultural context. The eKMC trial findings alone do not support a change to KMC policy, but a recent WHO multi-centre trial is influencing a shift towards immediate KMC. These findings provide rich data and insights into implementation and impact of SSNC, operationalisation of KMC for stable neonates, and novel data regarding the impact, feasibility, and realities of providing early KMC to unstable newborns in a typical African hospital neonatal unit.