The present paper reports and discusses three methodological considerations in research on cognitive aging that have theoretical consequences for the conclusions drawn in studies in the field and potentially for the development of future research in this area. The first issue is about cross-sectional data versus longitudinal data. It is argued that longitudinal data are to be preferred in studies of individual development and change. The second issue deals with the multidisciplinary nature of cognitive aging research. It is argued that such studies should involve behavioural data, brain imaging data and genetic data. For the third issue it is discussed that early cognitive data from childhood and genetic data might be regarded as a proxy for a hard-wired brain reserve that is interacting an experienced-based cognitive reserve that is developing and changing throughout adulthood and old age.