The aim of this investigation was to determine how cognitive performance was associated with positive and negative affect and life satisfaction over time. This study involved a secondary longitudinal analysis of cross-section data collected at Phase I (1988-1992) and during an 18-month longitudinal followup at Phase II (1992-1998) of the Georgia Centenarian Study. Participants included N = 137 centenarians at Time 1 and N = 68 survivors at Time 2. Significant stability in cognitive impairment existed at Time 1 and Time 2 for positive (β = .55, P < .01) and negative affect (β = .54, P < .01) models. Negative affect at Time 1 was associated with lower life satisfaction at Time 1 (β = -.42, P < .01 ). In addition, cognitive impairment at Time 2 was associated with decreased positive emotionality at Time 2 (β = -.39, P > .01). Furthermore, greater positive affect at Time 2 was associated with greater satisfaction with life at Time 2 (β = .35, P < .01). It appears that positive emotionality contemporaneously influences the association between cognitive impairment and life satisfaction among centenarians. Implications relative to improving life satisfaction among centenarians are discussed.