Background: Moderate alcohol consumption is associated with improved vascular risk profile and decreased mortality in the middle aged. An elevated homocysteine concentration is an independent risk factor for cardiovascular disease. Objective: To examine the relationship between alcohol consumption and homocysteine concentrations in severely obese patients (body mass index (BMI) > 35). Design: A careful alcohol history was obtained from 350 (male : female 1 : 5) consecutive patients as part of preoperative assessment for surgical treatment of obesity. Data were obtained concerning amount, frequency, timing and type of alcohol consumption. Fasting homocysteine, serum folate and vitamin B 12 concentrations were measured. Differences between groups were assessed using Student t-test, and ANOVA. Linear regression was used to assess factors influencing homocysteine concentration.Results: There is a U-shaped relationship between alcohol consumption and homocysteine concentrations, with light to moderate consumption being associated with lower concentrations. Those consuming < 100 g=week (n ¼ 165) of alcohol had geometric mean (95% CI of mean) serum homocysteine concentrations of 8.5 (8.2 -8.9) mmol=l compared with 9.5 (9.1 -9.9) mmol=l for non or rare consumers (n ¼ 153; P ¼ 0.001). The lower concentrations of homocysteine in regular consumers were associated with higher folate concentrations of 9.4 (8.6 -10.2) ng=ml when compared with non-consumers 7.5 (7.1 -7.8) ng=ml (P ¼ 0.001). Red wine consumers (n ¼ 42) had lower fasting concentrations of homocysteine 7.8 (7.5 -8.1) mmol=l compared with 153 non-consumers 9.4 (9.0 -9.8) mmol=l (P < 0.001), 82 beer and spirit consumers 9.0 (8.4 -9.7) mmol=l (P ¼ 0.005) and 73 white wine consumers 8.8 (8.2 -9.4) mmol=l (P ¼ 0.013). Red wine consumption was an independent predictor for lower homocysteine concentrations. Conclusion: Mild to moderate alcohol consumption, especially red wine consumption, in obese subjects is associated with lower fasting homocysteine concentrations. This may reduce cardiovascular risk and help explain the 'French paradox'.