Age-dependent diseases are those in which pathology is an intrinsic part of the ageing process. Age-related diseases, on the other hand, vary in incidence with age, but are not necessarily related to the ageing process. For an agedependent disease, incidence and mortality rise exponentially with age, while for an age-related disease the occurrence may increase with age and then decrease or may continue to increase in a less than exponential manner. This distinction may have important implications for prevention. The ultimate purpose of investigating the epidemiology of a disorder is to promote its prevention. In order to prevent a disorder it is desirable to understand its etiology. However, even without a full understanding of etiology, prevention can be carried out through modification of risk factors. This strategy is limited, however, by the fact that some risk factors are, by their nature, not modifiable. Furthermore, a risk factor is, strictly speaking, only a correlate of disease incidence. If the association between the risk factor and the disease is not a causal one, modification of the risk factor would have no effect. Because the etiology of neither Alzheimer's disease nor different dementias is understood, the strategy of modifying risk factors is presently the most viable option for prevention. In this review we discuss the situation if such a strategy was applied, what benefits might reasonably be expected. In answering, it is useful to consider first the degree to which chronic diseases occurring with age are preventable.