The ageing of the global population is the most important medical and social demographic problem worldwide. Promoting health among older adults, especially regarding nutrition and loneliness, can contribute to an increase in healthy life years and life expectancy. eHealth services can be used as a tool to promote health. However, they are often not tailored and persuasive enough to be effective in changing lifestyle behaviors. Because all individuals have different needs and preferences, it is essential to understand what works for whom and why. Only then can effective eHealth tools be offered that, combined, support the health and the well-being of people in older age. This dissertation focused on a specific tool: the Embodied Conversational Agent (ECA). The aim was to provide insight into how to design and evaluate ECAs that support healthy living. Through a series of studies (co-design sessions with older adults, qualitative experiments, and an evaluation), we identified the different factors that explain ECAs potential success for healthy ageing.