Emotional suffering in old age is largely caused by various psychiatric conditions, of which depression is the most common. Depression is associated with a decline in both well-being and daily functioning and reduces both morale and social capacity among the very old, which may produce high health and social costs for society. The overall aim of the thesis was to study the prevalence of depression among the very old, to identify factors associated with depression and to evaluate the prognosis of depression among the very old. In total, 363 people were evaluated for depression, 242 from an urban municipality in the year 2000 and 121 from five rural municipalities in 2002. In 2005, those still alive in the urban municipality were asked to participate again, and were therefore re-evaluated. The prevalence of depression was 27% in the urban municipality, 34% in the rural municipalities and 29% in the total sample. Of those depressed, about 67% were receiving antidepressive treatment, and of those, approximately 50% had responded to treatment. In the rural municipality, the depressed were less often treated with Selective Serotonin Re-uptake Inhibitor medications, receiving instead Tri-Cyclic Antidepressants. In the rural municipalities, only 38% of the depressed had responded to treatment. A higher proportion of women were diagnosed as depressed, 33% vs. 19%, p=0.006, although the response rate was the same for men and women. Depression was twice as common among those with dementia, 44% vs. 23%. There were discrepancies concerning associated factors between the depressed participants with dementia and those without. Experiencing the death of a child during the preceding ten years was associated with depression and independently associated with depression among men and participants with dementia. In all the studies, the depressed were less often able to go outside independently and to visit others. They also received fewer visits from others and often experienced loneliness. The great majority of those who were depressed in 2000 died during the subsequent five years, only 13 out of 65, 22%, were still alive in 2005, compared to 41% of those who were not depressed, p=0.003. Of 13 who survived, only two had recovered. Twenty-four out of 70 non-depressed people, 34%, had developed depression during the five years (2000-2005), and the total prevalence in year 2005 was 42% (35 out of 83 participants). Ten out of the 24 who had developed depression were prescribed antidepressants. Of those ten, four were regarded as responders. In the group with persistent depression, nine out of eleven were receiving antidepressants and 67% were responders. In conclusion, a large proportion of the very old suffer from under-diagnosed and undertreated depression. The response rate to treatment seems to be low, and the quality of treatment and follow-up also seems to be poor. The mortality rate among the depressed was high. The spectrum of factors associated with depression in people with dementia is different from that associated with depression among non-demented. Depression among the very old clearly emerges as a common and serious public health problem, with probably the most serious impact on quality of life. More efforts have to be made to improve the quality of assessments, treatment and research regarding depression among the very old.