There is currently little primary data looking at the epidemiology of back pain in older people and there are few large-scale population based studies which consider the occurrence or aetiology of back pain within this age group. Some evidence suggests that while non-disabling back pain decreases in the oldest old, the prevalence of disabling back pain may continue to increase. However, the aetiology of back pain in this group remains relatively unknown. Therefore, the overall aim of the current thesis was to investigate the epidemiology of back pain in older people, to examine the occurrence of back pain and age-related patterns in prevalence and incidence (descriptive epidemiology) and to identify potential risk factors for predicting back pain onset in those ≥75 years, specifically, to determine the role of social networks (analytical epidemiology). This was done through secondary analysis of a prospective cohort study from Cambridge looking at older people ≥75 years. It has been shown that while the prevalence of non-disabling back pain did not vary significantly across age, the prevalence of disabling back pain increased with age. In addition, it has been demonstrated, firstly, that aspects and indicators of physical health and a prior history of back pain are associated and independent predictors of back pain in older people; and secondly, that while objective measures of social contact are not risk markers for back pain, low mood, feelings of social isolation and depression are more strongly associated. These findings were further verified and confirmed in a second study, a cross-sectional survey of people ≥65 years living in rural Scotland. It can therefore be concluded that back pain is a common problem in the older population and, while mild back pain may remain constant across older age, disabling back pain continues to increase into the oldest old. The risk profile for back pain in older age is multifactorial, while some risk factors evident in younger populations remain; there is greater emphasis on the impact of mood and psychosocial factors.