The association between alcohol consumption and an increased risk of breast cancer has been established. It is still unclear however, whether this relationship differs across the estrogen receptor (ER) and progesterone receptor (PR) tumors subtypes. To provide a quantitative assessment of the association between alcohol intake and the risk of ER-/PR-defined breast cancer, we conducted a meta-analysis of cohort and case–control studies. Studies were identified by a literature search of PubMed through April 20, 2007 and by searching the reference lists of relevant articles. Summarized risk estimates (REs) with 95% confidence intervals (CIs) were calculated using random-effects models. The summarized results of the meta-analysis comparing the highest versus the lowest consumption categories showed statistically significant higher risks of developing all ER+ (27%), all ER− (14%), ER+PR+ (22%) and ER+PR− (28%), but not ER−PR− tumors. The dose–response meta-analysis showed that an increase in alcohol consumption of 10 g of ethanol per day was associated with statistically significant increased risks for all ER+ (12%), all ER− (7%), ER+PR+ (11%) and ER+PR− (15%), but not ER−PR−. A statistically significant heterogeneity of the REs across all ER+ versus ER−PR− was observed (pheterogeneity = 0.02). The summarized results from studies with adjustment for postmenopausal hormone use, body mass index and family history of breast cancer were higher and statistically significantly different from those without. The observed positive associations with alcohol for ER+PR+ and ER+PR− tumors cannot be explained by estrogen-dependent pathway only. Further studies need to clarify the biological mechanisms. © 2007 Wiley-Liss, Inc.