Abstract High age is a distinguished risk factor of falling, but there remains gaps in epidemiological data among very old people. Also, effects of exercise-oriented intervention implemented by geriatric teams is not known. The aims were to describe the incidence and risk factors of falls, the risk of fall-related injuries associated with physical activity, and to investigate the effect of pragmatic exercise intervention on fall risk among a home-dwelling population aged 85 years and older. The target population consisted of home-dwelling citizens of Oulu aged 85 years or more (N = 827). Altogether 555 people (mean age 88 SD ± 3), representing 67% of the population sample, were evaluated by interviews and clinical assessments. Falls and physical activity were monitored via telephone nine times during a 27-month follow-up constituting 1114 person years. Risk factors of falls were determined during an 11-month period before intervention, as were the effects of the intervention during a 16-month period thereafter. Negative binomial regression, pooled logistic regression and Cox regression analyses were used to analyze data. The incidence rates of falls, major soft tissue injuries and fractures were 1039 (95% CI 974–1093), 74 (58–92) and 89 (72–108), respectively. The probability of getting injured was higher in the morning and evening than in the daytime. The contribution of the ongoing activity and the type of falling to the risk of injury was less than that previously reported among younger home-dwellers. The risk factors of falls were a history of recurrent falling, trouble with vision when moving, use of an antipsychotic drug, and feelings of anxiety, nervousness or fear. Exercise other than walking was associated with a reduced risk of injury-causing falls. Pragmatic intervention wasn’t effective in preventing falls, but it was effective in preserving balance performance. Among those with better functional abilities, intervention was effective in reducing the risk of the first four falls. Adherence to recommended interventions was relatively low. In conclusion, the frequency of falls and fall-related fractures increases up to the highest ages. Anxiety-related disorders may be more important risk factors of falls than are drugs commonly used in treatment. Exercise related to everyday activities is safe among the most elderly. The effects of practical exercise intervention are promising, but attention needs to be paid to adherence to exercise in order to improve these effects.