Background— This study tested the hypothesis that moderate alcohol intake exerts its cardioprotective effect mainly through an increase in the serum level of high-density lipoprotein cholesterol. Methods and Results— In the Cohort of Norway (CONOR) study, 149 729 adult participants, recruited from 1994 to 2003, were followed by linkage to the Cause of Death Registry until 2006. At recruitment, questionnaire data on alcohol intake were collected, and the concentration of high-density lipoprotein cholesterol in serum was measured. Using Cox regression, we found that the adjusted hazard ratio for men for dying from coronary heart disease was 0.52 (95% confidence interval, 0.39–0.69) when consuming alcohol more than once a week compared with never or rarely. The ratio changed only slightly, to 0.55 (0.41–0.73), after the regression model included the serum level of high-density cholesterol. For women, the corresponding hazard ratios were 0.62 (0.32–1.23) and 0.68 (0.34–1.34), respectively. Conclusions— Alcohol intake is related to a reduced risk of death from coronary heart disease in the follow-up of a large, population-based Norwegian cohort study with extensive control for confounding factors. Our findings suggest that the serum level of high-density cholesterol is not an important intermediate variable in the possible causal pathway between moderate alcohol intake and coronary heart disease.