This study investigates the characteristics of and the interrelationships between disease, disability, service use and social support in a random sample of 647 community dwellers aged 75 years and over. The two broad objectives of the study are: to examine the physical aspects and manifestations of health by investigating disease and disability and the interrelationships between these two factors, and; to examine the social aspects of health by investigating service use and social support and the interrelationships between these two factors. Given the dramatic population ageing in Australia, particularly in the very old age groups, the health, well-being and quality of life of older Australians are of paramount importance and will be well into the future. The proportion of the population with diseases and disabilities increases significantly with age. As the physical aspects of health are manifested with increasing age the social aspects of health also become increasingly important. Older adults, particularly those in advanced old age, are disproportionately high users of health and community services. Despite the high use of services in this age group, far more older adults living in the community rely on their families, friends and neighbours for social support and many older adults use a combination of formal services and informal social support. Little is known about people aged 75 years and over living in the community in Australia. In particular, significant knowledge gaps exist with regard to the relationship between disease and disability and that between service use and social support. The characteristics of social support in this group of older adults are also largely unknown. The papers presented in this thesis are based on data collected in The Sydney Older Persons Study (SOPS). This is a large longitudinal multidisciplinary project which began in 1991 in order to investigate the health and service use patterns of people aged 75 years and over living in the community in the Central Sydney Health Area. The initial sample consisted of two groups: first, the Australian Bureau of Statistics (ABS) selected census districts with probability proportional to size and 9271 households were door-knocked to obtain a random sub-sample of the general community (n=320, response rate 73%); second, community-living veterans and war widows residing in the Central Sydney Health Area were selected at random from a list provided by the Department of Veterans Affairs to obtain a veteran/war widow sub-sample (n=327, response rate 82%). Respondents participated in both an interview conducted by a social scientist and a medical assessment performed by a medical practitioner with experience in geriatric medicine. An informant was sought for each respondent and this informant participated in a phone interview conducted by a social scientist. The first paper in this thesis investigates the characteristics of diseases (neurodegenerative, systemic and psychiatric) including their prevalence and association with age. The second paper extends the first by examining the nature of the relationship between disease and disability and in particular which individual diseases and groups of diseases have the greatest impact on disability. The third paper expands the analysis in the second paper by focusing in greater detail on the relationship between disease and disability. The contribution of clinically-diagnosed individual diseases and groups of diseases to three different measures of disability (clinician-rated, informant-rated or proxy and self-report) is investigated here. The fourth paper examines the possibility of disease and disability being the major predictors of service use and social support. It focuses on the determinants of service use and social support using Andersen's behavioral model. The fifth paper investigates the characteristics of social support, in particular gender differences and the socio-demographic variables associated with social support. This is an important research area because lower levels of social support have been found to predict mortality, disease and lower levels of well-being. Finally, the sixth paper links the major themes of the fourth and fifth papers by investigating the relationship between service use and social support. This paper tests Cantor's 'hierarchical-compensatory' mechanism, which predicts a negative association between service use and social support, and the 'bridging' mechanism which predicts a positive association between these two factors. Thus it assesses the extent to which demands for service use and for social support are made together or in a compensatory fashion for respondents of equal disease and disability. The presented work demonstrates that neurodegenerative diseases [dementia, cognitive impairment, parkinsonism, instability (gait ataxia), immobility (gait slowing) and motivation loss/behaviour change] have the largest and most significant increases with age of all disease groups. Therefore the hypothesis made in paper one that neurodegenerative diseases will come to dominate the health care needs of older adults, particularly when combined with population ageing, is supported. Further, results of papers two and three indicate that neurodegenerative diseases result in greater levels of disability, lending credence to the finding that it is these neurodegenerative diseases that are of central importance to the future of the health care needs of older adults of advanced age. While systemic diseases play an important role in disability, the neurodegenerative diseases are under-recognised by self-report and yet are most strongly associated with severe disability. A major recommendation of this study is that assessments and diagnosis of neurodegenerative diseases be included in disability assessments. With regard to the social aspects of health, the fourth paper finds that disease and disability are the main predictors of service use and social support. The fifth paper highlights important gender differences in social support and also finds that lower levels of social support are associated with increased age, male gender, single marital status and lower socioeconomic status. Because it is widely accepted that social support is protective against adverse health outcomes and low levels of wellbeing, these groups of older adults are at risk of poorer health and wellbeing. Finally the sixth paper fills some knowledge gaps with regard to the relationship between service use and social support. It shows that with regard to IADL (instrumental activities of daily living) services and IADL social support, Cantor's 'hierarchical-compensatory' mechanism (negative correlation) applies but with regard to medical services and both ADL (activities of daily living) and IADL social support the 'bridging' mechanism (positive correlation) is supported. These complex interrelationships between disease, disability, service use and social support are summarised schematically in a model. In light of significant population ageing, substantial resources in the form of medical and community services and social support from carers, family, friends and neighbours will need to be devoted to older adults with diseases, in particular neurodegenerative diseases, and to those with disabilities. Given the increasing importance of disease, disability, service use and social support in very old age, it is crucial that knowledge and understanding of these factors and their interrelationships be advanced in order to better allocate and sustain resources and to ultimately improve the health, well-being and quality of life of very old adults.