BACKGROUND: The deleterious health effects of sedentary behaviors, independent of physical activity, are increasingly being recognized. However, associations with cognitive performance are not known. PURPOSE: To estimate the associations between different sedentary behaviors and cognitive performance in healthy older adults. METHODS: Computer use, time spent watching television (TV), time spent reading and habitual physical activity levels were self-reported twice (in 2001 and 2007) by participants in the SUpplémentation en Vitamines et MinérauX (SU.VI.MAX and SU.VI.MAX2) study. Cognitive performance was assessed at follow-up (in 2007-2009) via a battery of 6 neuropsychological tests used to derive verbal memory and executive functioning scores. Analyses (ANCOVA) were performed among 1425 men and 1154 women aged 65.6 ± 4.5 at the time of the neuropsychological evaluation. We estimated mean differences with 95% confidence intervals (95%CI) in cognitive performance across categories of each type of sedentary behavior. RESULTS: In multivariable cross-sectional models, compared to non-users, participants using the computer for \textgreater1 h/day displayed better verbal memory (mean difference=1.86; 95%CI: 0.95, 2.77) and executive functioning (mean difference=2.15; 95%CI: 1.22, 3.08). A negative association was also observed between TV viewing and executive functioning. Additionally, participants who increased their computer use by more than 30 min between 2001 and 2007 showed better performance on both verbal memory (mean difference=1.41; 95%CI: 0.55, 2.27) and executive functioning (mean difference=1.41; 95%CI: 0.53, 2.28) compared to those who decreased their computer use during that period. CONCLUSION: Specific sedentary behaviors are differentially associated with cognitive performance. In contrast to TV viewing, regular computer use may help maintain cognitive function during the aging process. CLINICAL TRIAL REGISTRATION: clinicaltrial.gov (number NCT00272428).