Background: With increasing age, older adults experience declines in health, often in the form of reduced functionality, increased multimorbidity and waning cognitive capacity. Despite such declines, some older adults provide stable (and sometimes even improving) subjective appraisals of their health over time – even in the face of declining objective health. Growing dissociations between subjective and objective health scores in later life mean that some older adults may provide overly ’health optimistic’ or ’health pessimistic’ appraisals of their health. However, consensus on how these individuals should be identified is lacking. Additionally, the long-term health implications of being overly health optimistic or pessimistic in later life is considerably underexplored. Methods: To address this, the present thesis describes a health asymmetry framework, a metric which classifies older adults into groups, based on the level of agreement between their subjective and objective health scores (‘health optimistic’, ‘health pessimistic’, ‘good health realistic’ and ‘poor health realistic’). Using four nationally representative, secondary archived datasets across Europe, including one instance of primary data collection, the thesis longitudinally investigates how these health asymmetry categories are associated with mental and functional health sequelae over time. STUDY 1 utilises data from wave one of the Irish Longitudinal Study of Ageing to create a health asymmetry metric. Self-rated health scores were compared to scores from a Frailty Index, in a sample of n=6907 older adults (aged 50+ years), resulting in the derivation of the following categories: ’health optimistic’, ’health pessimistic’ and ’health realistic’. Multinomial logistic regressions found that being health pessimistic was associated with psychosocial factors such as increased anxiety, increased loneliness and decreased social connectedness. In contrast, being health optimistic was associated with increased levels of vigorous exercise and alcohol consumption. STUDY 2 describes how a sample of older English adults transitioned from one health asymmetry category to another, across waves one to three of the English Longitudinal Study of Ageing. Using first-order Markov transition and generalised logit models, health realistic individuals were likely to remain health realistic over time. The prevalence of the health optimistic category increased over two wave transitions, while the health pessimistic category yielded volatile, unstable transition probabilities. STUDY 3 investigated how baseline health asymmetry status predicted depressive symptoms and change in depressive symptoms over time in European older adults. Using data from the Survey of Healthy Ageing and Retirement in Europe, multilevel growth curve models found that health pessimists consistently had the highest levels of depressive symptoms over time. Health optimists had a gradually declining trajectory of depressive symptoms across a 14-year study period. STUDY 4 explored whether health asymmetry categories differentially predicted the risk of an injurious fall in older Swedish adults. Data from the Swedish National Study on Ageing and Care in Kungsholmen were utilised. A set of time-varying Cox and Laplace regressions found that health optimists had the most elevated risk of experiencing an injurious fall of all health asymmetry categories, which may be due to their poor OH and optimistic biases which make health pessimists believe they are not at risk of falling. STUDY 5 involved intensive primary data collection from a sample of n=53 middle aged and older Irish adults, through a smartphone-based Ecological Momentary Assessment design. Health anxiety data were collected over the course of a 6-day study period. Growth mixture modelling found that two latent trajectories of health anxiety evolved over the study period. Health asymmetry categories did not significantly predict health anxiety scores over time. Conclusions: Health asymmetry appears to be a clinically relevant tool which is significantly associated with depressive symptomatology, injurious falls and mortality. Health asymmetry may be more useful as a summary tool for large-scale population-based studies, given that the metric can be retrospectively constructed using secondary data. Future research should consider exploring the association between health asymmetry and other important physical and functional health outcomes in older adults, such as psychological distress and multimorbidity.