The prevention of neurodegenerative dementias, such as Alzheimer's disease, is a public health priority. Due to the large numbers of affected patients, even interventions bringing about a relatively small delay in disease onset could have large public health effects. Randomized controlled trials (RCTs) are required to demonstrate the effectiveness of preventive interventions, but such trials raise specific methodological questions because they are new in the field of neurodegenerative diseases, and require large numbers of elderly subjects and lengthy follow-up periods. We performed a literature search to identify primary prevention RCTs for neurodegenerative dementia. The methodology of the trials was summarized and discussed during two expert meetings. Overall, 39 trials were identified that assessed dementia incidence or cognitive decline as a primary or secondary study outcome. Age was the most common selection criteria for target populations. Follow-up periods ranged from one month to nine years and were longest in studies measuring dementia incidence as an outcome. Results of RCTs have so far been generally negative and conflicting with those of observational studies, perhaps due to methodological issues. Future trials must therefore carefully consider the target population, outcomes and duration of follow-up to be used, and should assess the problem of attrition.