The authors evaluated the association of low-to-moderate alcohol consumption with risk of cognitive decline in a census-based cohort study of men and women aged ≥55 years conducted in Zaragoza, Spain (1994-1999). Participants free of dementia at baseline (N = 3,888) were examined after 2.5 and 4.5 years of follow-up. Information on alcohol intake was collected with the EURODEM Risk Factors Questionnaire and the History and Aetiology Schedule. The study endpoint was severe cognitive decline, defined as loss of ≥1 point/year on the Mini-Mental State Examination or a diagnosis of incident dementia (Diagnostic and Statistical Manual of Mental Disorders: DSM-IV, Text Revision criteria). Compared with those for abstainers, the multivariate-adjusted odds ratios for severe cognitive decline for male drinkers of <12 g alcohol/day, drinkers of 12-24 g alcohol/day, and former drinkers were 0.61 (95% confidence interval (CI): 0.31, 1.20), 1.19 (95% CI: 0.61, 2.32), and 1.03 (95% CI: 0.59, 1.82), respectively. The corresponding odds ratios for women were 0.88 (95% CI: 0.45, 1.72), 2.38 (95% CI: 0.98, 5.77), and 1.03 (95% CI: 0.48, 2.23). This study did not support the hypothesis that low-to-moderate alcohol consumption prevents cognitive decline. The inverse association between low-to-moderate alcohol intake and cognitive decline observed in other studies may have been due to inclusion of former drinkers in the abstainers reference category.