The impact of total serum cholesterol as a risk factor for cardiovascular disease decreases with age, which casts doubt on the necessity for cholesterol-lowering therapy in the elderly. We assessed the influence of total cholesterol concentrations on specific and all-cause mortality in people aged 85 years and over. In 724 participants (median age 89 years), total cholesterol concentrations were measured and mortality risks calculated over 10 years of follow-up. Three categories of total cholesterol concentrations were defined: <5·0 mmol/L, 5·0–6·4 mmol/L, and ⩾6·5 mmol/L. In a subgroup of 137 participants, total cholesterol was measured again after 5 years of follow-up. Mortality risks for the three categories of total cholesterol concentrations were estimated with a Cox proportional-hazards model, adjusted for age, sex, and cardiovascular risk factors. The primary causes of death were coded according to the International Classification of Diseases (ICD-9). During 10 years of follow-up from Dec 1, 1986, to Oct 1, 1996, a total of 642 participants died. Each 1 mmol/L increase in total cholesterol corresponded to a 15% decrease in mortality (risk ratio 0–85 95% Cl 0·79–0·91). This risk estimate was similar in the subgroup of participants who had stable cholesterol concentrations over a 5-year period. The main cause of death was cardiovascular disease with a similar mortality risk in the three total cholesterol categories. Mortality from cancer and infection was significantly lower among the participants in the highest total cholesterol category than in the other categories, which largely explained the lower all-cause mortality in this category. In people older than 85 years, high total cholesterol concentrations are associated with longevity owing to lower mortality from cancer and infection. The effects of cholesterol-lowering therapy have yet to be assessed.