The effects of alcohol consumption and tobacco smoking on the prevalence of esophageal cancer vary considerably by country, race and lifestyle. Few data exist on the effect of the interaction between the amount and duration of alcohol consumption and tobacco smoking on the incidence of esophageal cancer. In this case-control study, the cases included patients with histologically confirmed esophageal squamous cell carcinoma (ESCC) younger than 60 years of age and recruited between January 1, 2002 and December 31, 2006. The controls had no abnormality during a medical checkup. A total of 835 pairs were created by pairing each case to a gender- and age-matched control. Conditional logistic regression analysis was used to estimate adjusted odds ratios (ORs) and 95% confidence intervals. Univariate conditional logistic regression analyses revealed that the ORs according to both duration of alcohol consumption and tobacco smoking increased monotonically. Alcohol consumption and tobacco smoking may have a synergistic effect on the incidence of ESCC. Conditional logistic regression analysis using a forward stepwise selection procedure revealed that the incidence of ESCC was associated with the duration of tobacco smoking, the interaction between the amount and duration of alcohol consumption, and a family history of cancer. In particular, groups with a long duration of alcohol consumption and high alcohol intake had much higher ORs than those with short duration and low intake, which highlights the importance of the interaction between the amount and duration of alcohol intake. This study confirmed the significance of the interaction between alcohol consumption and tobacco smoking in esophageal cancer. This interaction between amount and duration is an accurate indicator for estimating the risk of esophageal cancer attributable to alcohol consumption and tobacco smoking. These findings suggest that decreasing the number of young and middle-aged drinkers and smokers will reduce the incidence of esophageal cancer.