OBJECTIVE: To examine the risk of CHD in relation to alcohol intake from three different instruments. DESIGN: In the European Prospective Investigation into Cancer in Norfolk study, weekly alcohol intake was estimated from a single question in a mail-in health and lifestyle questionnaire (HLQ), a semi-quantitative FFQ, and a 7 d diet diary (7DD). Information on smoking status, physical activity, disease history, social class and medication use was reported in the HLQ. Height, weight, blood pressure and blood lipids were measured at a health check-up. The average length of follow-up was 11 years. The association between alcohol intake and incident fatal and non-fatal CHD in a nested case-control sample was calculated using logistic regression. SETTING: Norfolk, England. SUBJECTS: A total of 2151 cases of incident fatal and non-fatal CHD and 5354 controls. RESULTS: The Spearman correlation values between the 7DD, FFQ and HLQ alcohol estimates ranged from r = 0.70 to 0.82 (P < 0.0001 for all r values). Alcohol intake from all instruments was inversely associated with the risk of CHD in age- and multivariate-adjusted models. The relationships between the risk of CHD and alcohol intake from the 7DD, HLQ or FFQ were not significantly different from each other (P >0.10). A marginal difference between men and women was detected for the risk of CHD in relation to HLQ alcohol intake (P = 0.065). CONCLUSIONS: In conclusion, while the instruments were not uniform in their assessment of alcohol intake levels, the 7DD, HLQ and FFQ yielded similar inverse associations between alcohol intake and risk of CHD.