Objective: To investigate the effects of light to moderate alcohol consumption on cognitive performance. Design and Setting: A cross-sectional analysis including older Japanese Americans in King County, WA, enrolled in the Kame Project, a population-based study of cognition, dementia, and aging. Participants: 1,836 cognitively intact participants aged 65 and older who participated in the baseline (1992-1994) examination. Measurement: Cognitive performance was measured using the Cognitive Abilities Screening Instrument, reaction time (simple and choice), and a measure of vocabulary (North American Adult Reading Test). Results: Multivariate analyses were used to examine the relationship between cognitive performance and alcohol consumption at baseline with men and women together and then separately controlling for age, education, smoking, history of stroke, angina, hypertension, diabetes, and coronary heart disease. Findings showed lower cognitive test scores were observed for men who were either abstainers or in the heavy drinking group. For women, a linear relationship between alcohol consumption and cognitive performance was seen on two of the four measures of cognitive functioning. No significant difference in the association of drinking and cognitive function was identified within the different Japanese American subgroups. Conclusion: Results suggest a possible positive relationship between light to moderate drinking and cognitive performance in an aging Japanese American population. Additional long-term prospective and cross-cultural studies are needed to determine the generalizability of these findings to other aging cohorts.