Childhood malnutrition is a growing public health challenge for low- and middle-income countries. Globalization and urbanization have led to a significant shift in dietary habits and physical activity patterns over the past few decades, resulting in the coexistence of undernutrition and overnutrition within populations. The negative health, development, and social impacts of childhood malnutrition are serious and ongoing for individuals and their families, while also contributing an economic burden to these countries. Early childhood is a crucial period in the formation of dietary habits, which tend to track throughout life. Worldwide, current evidence suggests that the dietary habits of preschool-aged children, namely, low intake of fruit and vegetables, and higher intake of foods rich in sugar, fat, and salt, do not align with dietary recommendations. Therefore, early interventions are important to establish healthy dietary habits in preschool children to combat childhood malnutrition. Sri Lanka is a middle-income country that is currently experiencing a simultaneous rise in overweight and obesity along with continued undernutrition in children aged five years or younger. This is more pronounced in urban locations of Sri Lanka, which comprises approximately one third of all children aged five years or younger. To date, multiple strategies have been implemented to improve the nutritional status of Sri Lankan children; however, none have specifically targeted preschool-age children. Therefore, population-specific effective eating interventions are urgently warranted for urban Sri Lankan preschool children. Development of food choices is multifactorial; hence, using an ecological approach to understand the interplay of these factors in the formation of food choices in early childhood is necessary. Addressing multi-level factors that influence children’s food choices has improved the effectiveness of intervention strategies. However, such an understanding and subsequent application of intervention strategies have not been examined in Sri Lanka. Therefore, this research investigated the application of an ecological approach to improve healthy eating in urban preschool children (aged 2-6 years) in Sri Lanka. This thesis incorporates three research studies designed to address the overall research aim. Each of the three studies informs subsequent studies in the research program to enable a deeper understanding of ecological factors shaping preschool children’s food choices and the effectiveness of an eating intervention strategy which applies an ecological approach to improve food knowledge, preferences, and choices. The findings have been published or are under review in peer-reviewed journals. This research was underpinned by a pragmatic paradigm to accommodate the use of a mixed methods approach, as it enabled a rich understanding of the socioecology of children’s food choices, to promote healthy food choices in children. A conceptual framework was developed to guide this research. Study 1 was a systematic review of the literature that examined interventions targeting the family and community factors influencing the eating behaviour of preschool children in low- and middle-income countries. Thirteen intervention studies which met the inclusion criteria were critically appraised. The review highlighted that future interventions should focus on improving the nutritional knowledge of the family or caregivers while creating a supportive family environment via increasing household food availability and family income to promote healthy eating behaviours in low- and middle-income country preschool children. Study 2 was a qualitative exploration with urban Sri Lankan parents/caregivers to understand their perceptions of factors that influence their preschool child’s food choices. Thirteen focus groups, adopting a partial nominal group technique, were conducted with parents/caregivers. A quantitative approach was involved to select and rank influential factors. Qualitative data were summarised using content analysis. The findings showed that parents/caregivers perceived a child’s food preferences was the core driving factor in their food decisions. Parents’ and caregivers’ nutritional knowledge, maternal control of food choices, family income, and household food preparation facilities were the next most influential factors. Study 3 included a baseline cross-sectional assessment and a cluster randomised controlled trial. The cross-sectional survey was conducted to understand the dietary habits of Sri Lankan urban preschool children, which assessed their dietary diversity and dietary intake. Average daily dietary intakes for selected food and beverage groups were computed in servings and frequencies descriptively. Findings demonstrated that dietary diversity of children was within the medium category, but their dietary intakes were not aligned with the national recommendations, revealing relatively unhealthy dietary habits. More than 80% of preschool children in this cohort consumed an inadequate daily intake of fruit and vegetables, and one in ten had sugary snacks and confectionary at least twice a day. In addition, the unhealthy dietary and social behaviour of watching television while eating the evening meal was reported in more than one third of children. The cluster randomised controlled trial involved a multicomponent eating intervention (MCI) aimed to improve the food knowledge, healthy food preferences and dietary diversity in urban Sri Lankan preschool children. The trial was conducted for six weeks with two additional intervention arms: single component intervention (SCI) and control (no intervention). As per the ecological approach, the MCI incorporated family engagement which included: parental nutrition education; engaging children in meal preparation; and tasting exposure for children with parent, teacher and peers. The effectiveness of MCI over SCI or control was determined using difference-in-difference model. The findings revealed that compared to the control arm, both MCI and SCI significantly increased children’s food knowledge and healthy food preferences, but not the dietary diversity score. MCI showed greater effect over SCI, reaching significance for children’s food knowledge. A delayed post-intervention assessment was suggested to evidence improved sustained effects of MCI on children’s healthy food preferences and dietary diversity scores, as changes in food preferences and eating behaviours may further develop with time. Collectively, the findings of this research program highlight the importance of an ecological perspective to understand how food choices are formed and how an ecological approach can be applied. The findings from this research can inform parents, families, preschool centre staff, and policy makers of the importance of applying an ecological approach to understand the contextual factors that shape children’s food choices and to develop strategies to promote healthy eating in preschool children. Further, Sri Lanka’s national policy on Early Childhood Development should incorporate parental and family engagement in preschool centre based health and nutrition activities. Embedding children’s nutrition education in the preschool education curriculum should be considered by Early Childhood Development/preschool curriculum developers. As the food choices formed during early childhood invariably continue throughout life, actions to establish healthy eating behaviours in this early life stage may help overcome childhood malnutrition and support a lifetime of healthy food choices.