Backgrounds Views on the relationship between alcohol consumption and stroke risk remain controversial. Moreover, data on cumulative alcohol intake are limited. We examined the potential impact of cumulative alcohol consumption on the risk of total stroke and its subtypes in men. Methods This prospective study included 23,433 men from the Kailuan Study. Cumulative alcohol consumption was taken as the primary exposure by calculating self-reported alcohol consumption from three consecutive examinations (in 2006, 2008, and 2010). The first occurrence of stroke was confirmed by reviewing medical records from 2010 to 2016. We used Cox proportional hazards regression to analyze the data. Results During the 5.9 +/- 0.8 years of follow-up, 678 total strokes were identified, including 595 ischemic stroke (IS), 90 intracerebral hemorrhage and 19 subarachnoid hemorrhage cases. The adjusted hazard ratios (95% confidence intervals) of total stroke for light, moderate and heavy cumulative alcohol consumption were 1.23 (1.01-1.51), 1.49 (1.13-1.97), and 1.50 (1.21-1.86), respectively, compared with those of nondrinkers. The results were similar for IS. Cumulative alcohol consumption was not associated with intracerebral hemorrhage risk (hazard ratio 1.46; 95% confidence interval, 0.74-2.08). Conclusions Cumulative alcohol consumption is an independent risk factor of total stroke and IS in men in a community-based cohort. Even light alcohol intake increases the risk of total stroke and IS.