BACKGROUND AND OBJECTIVE Changes in lipoproteins and hemostasis only incompletely explain the reduced cardiovascular mortality associated with light to moderate alcohol consumption. Since increasing evidence suggests that atherosclerosis can be considered to be a chronic inflammatory process, we sought to assess the association between daily alcohol consumption and levels of sensitive markers of inflammation. STUDY PARTICIPANTS AND METHODS 478 voluntary blood donors (358 men, 120 women) aged 40 to 68 years were categorized into four groups according to their self-reported amount of daily alcohol consumption: 0 g/day, >0 - 20 g/day, >20 - 40 g/day, and > 40 g/day. Means of various sensitive markers of inflammation (C-reactive protein (CRP), serum amyloid A (SAA), interleukin-6 (IL-6), intercellular adhesion molecule-1, plasma viscosity und albumin) were calculated and compared by bivariate and multivariate analyses. RESULTS More than 80 % of the study participants reported to consume alcohol, mainly beer. We found statistically significantly decreased levels of SAA, CRP, and plasma viscosity in subjects with light-to-moderate alcohol intake (>0 - 20 g/day and > 20 - 40 g/day, respectively), and a trend for increased levels of albumin in these subjects compared to non-drinkers. After multivariable adjustment for potential confounders (age, gender, body mass index, cigarette smoking, years of school education, and physical activity) a significant U-shaped association (p = 0.02) between levels of SAA and the amount of daily alcohol intake remained: there were 0.75 mg/l and 0.70 mg/l lower mean levels, respectively, of SAA in subjects with light-to-moderate alcohol intake compared to those of non-drinkers. Subjects with an alcohol intake of > 40 grams per day showed a statistically significant increase in levels of interleukin-6 (0.50 pg/ml) compared to non-drinkers. CONCLUSION Potential anti-inflammatory properties of moderate alcohol consumption might represent an additional mechanism to explain its atheroprotective effect.