Context: Reduced testosterone levels have been implicated as a potential causative factor in cognitive decline with older age. Men who possess the apolipoprotein E (APOE) ε4 allele have an increased risk of developing Alzheimer's disease; however, no studies have examined whether the influence of testosterone on cognition in healthy older men may be modulated by this genetic predisposition. Objective: The objective of the study was to investigate the association between serum testosterone concentrations and cognitive performance in healthy older men, taking into account APOE ε4 status. Design: This was a cross-sectional study conducted from 2003 to 2004. Setting: The study population consisted of community-dwelling males residing in Perth, Western Australia. Participants: Healthy men over 55 yr, free of cognitive impairment and dementia (n = 45), were included in the study. Main Outcome Measures: Participants had fasting early morning blood samples for testosterone and SHBG and were assessed for mood as well as indices of general cognition, verbal and visual memory, executive functioning, working memory, and attention. Results: There was a significant interaction between calculated free testosterone (FT) and APOE ε4 on general cognition (P = 0.01) and executive functioning, working memory, and attention (P < 0.01). Higher levels of FT were associated with better general cognition in non-ε4 carriers (P = 0.01). By contrast, in ε4 carriers higher FT levels were associated with lower scores on tests of executive functioning, working memory, and attention (P = 0.02). In men at increased risk for Alzheimer's disease, higher testosterone levels were not associated with better cognitive function. Conclusions: Cross-sectional and prospective studies of testosterone and cognition in older men should take into account APOE ε4 status.