OBJECTIVES: The effect of light to moderate alcohol consumption on the liver is controversial. To determine the association between light to moderate alcohol consumption and frequency of hypertransaminasemia, a cross-sectional and a subsequent longitudinal cohort study were conducted using annual health checkup data at a Japanese workplace. METHODS: We analyzed 1,177 male subjects (age 20-59) without HCV or HBV infection or other chronic liver diseases. To determine the association between alcohol consumption (none or minimal <70 g/wk, light > or =70 g and <140 g/wk, moderate > or =140 g and <280 g/wk, excessive > or =280 g/wk) and hypertransaminasemia, we performed multiple logistic regressions. We then followed 326 subjects without a history of fatty liver or hypertransaminasemia up to 5 years for incidental hypertransaminasemia and performed Cox proportional hazard regressions. RESULTS: Excess alcohol consumption was associated with increased odds of hypertransaminasemia (adjusted odds ratio [AOR]versus none or minimal consumption 1.4[1.1-1.93], P= 0.023). There was significant interaction between age group and alcohol consumption (P < 0.01). In the younger group, moderate consumption was associated with decreased odds (AOR 0.5 [0.3-0.9], P= 0.032), while in the older group, light consumption was associated with decreased odds (AOR 0.6 [0.4-1.0], P= 0.036) and excess consumption was associated with increased odds (AOR 1.6 [1.1-2.3], P= 0.014) of hypertransaminasemia. During follow-up, moderate consumption was associated with decreased incidence of hypertransaminasemia versus none or minimal consumption (adjusted hazard ratio 0.4 [0.1-0.9], P= 0.02). CONCLUSION: Light to moderate alcohol consumption may protect against the development of hypertransaminasemia among male subjects without other liver conditions. Further studies are required before recommending light to moderate alcohol consumption.