In this chapter, Moffat reviews observational research in older men that examines serum testosterone concentrations in relation to cognitive aging or Alzheimer's disease (AD) risk. Prospective cohort studies, such as the Baltimore Longitudinal Study of Aging, provide particularly useful data. Despite methodological limitations and conflicting findings, Moffat tentatively concludes that age-associated reductions in testosterone concentrations are a risk factor for cognitive decline and dementia. Cognitive vulnerability may be limited to specific domains of cognitive performance and may be modified by apolipoprotein E genotype. As he suggests, more definitive answers may require well designed randomized clinical trials that target cognitive effects of testosterone therapy. Any future trial must also consider other health outcomes that may be beneficially or adversely affected by testosterone.