The aim of this study was to evaluate the effects of anti-inflammatory intake on cognitive function in 7234 community-dwelling elderly persons. Cognitive performance, clinical diagnosis of dementia, and anti-inflammatory use were evaluated at baseline, and 2, 4, and 7 years later. Multivariate logistic regression analyses were adjusted for sociodemographic, behavioral, physical, mental health variables, and genetic vulnerability (apolipoprotein E ε4). Elderly women taking inhaled corticosteroids were at increased risk for cognitive decline over 7 years in executive functioning (odds ratio, 1.76; 95% confidence interval, 1.14-2.71; p = 0.04); the effect being increased after continuous use (odds ratio, 3.15; 95% confidence interval, 1.29-7.68; p = 0.01) and not found after discontinuation of treatment. In men, no significant associations were observed. Corticosteroid use was not significantly associated with an increase risk of incident dementia over 7 years. Nonsteroidal anti-inflammatory drug use was not significantly associated with either dementia incidence or cognitive decline in both sexes. The association may be related to hypothalamic-pituitary-adrenal corticotropic axis dysfunctioning rather than a direct anti-inflammatory mechanism. Long-term use of inhaled corticosteroids may constitute a form of reversible cognitive disorder in elderly women. Physicians should check this possibility before assuming neurodegenerative changes.