Abstract Background The Global Burden of Disease and Risk Factor models assess alcohol to be a leading cause of premature death both globally and in Mexico, but direct epidemiological evidence from Mexico is limited. We investigated the relevance of alcohol to mortality at ages 35-74 in a large prospective study of Mexican adults followed for nearly two decades. Methods Between 1998 and 2004, 150,000 adults aged ≥35 years were recruited into the Mexico City Prospective Study, and followed for median 19 years. Cox regression was used to relate baseline-reported alcohol consumption (never, former, ‘occasional’ [defined as less than monthly], and ‘regular’ [defined as at least monthly, and split into <70, ≥70 to <140, ≥140 to <210 and ≥210g/week]) to mortality at ages 35-74 from all causes, and from underlying causes selected a priori as being potentially ‘alcohol-related’. Analyses were adjusted for age at risk, sex, residential district, education, smoking, physical activity and diabetes. They excluded those who at recruitment were aged ≥75 years, had incomplete data, or had vascular disease or any other chronic condition (except diabetes) recorded. Results Among 45,262 men and 93,607 women aged 35-74 years, 6% of men and 26% of women reported being never drinkers, 45% of men and 56% of women reported being occasional drinkers, and 33% of men and 7% of women reported being regular drinkers. There were 12,597 deaths from any cause before age 75, including 3,433 deaths from the selected underlying causes. Compared with occasional drinkers, never drinkers had 16% higher all-cause mortality (RR 1.16, 95% CI 1.10-1.22) but similar mortality from the subset of deaths considered potentially alcohol-related (RR 1.06, 95% CI 0.96-1.17). Among regular drinkers, all-cause mortality increased progressively across the four studied categories. Compared with occasional drinkers, regular drinkers of ≥140 to <210g/week had 20% higher mortality (RR 1.20, 95% CI 1.08-1.34) while regular drinkers of >210g/week had 46% higher mortality (RR 1.46, 95% CI 1.33-1.61). For the subset of deaths considered to be alcohol-related these RRs were 1.61 (1.35-1.93) and 2.42 (2.10-2.79) respectively. Liver disease was especially strongly related to alcohol consumption: the RR comparing regular drinkers of >210g/week with occasional drinkers was 5.89 (95% CI 4.75-7.29). Associations were similar in those with and without diabetes, so the absolute relevance of alcohol to mortality was larger in those with diabetes. Conclusions In this Mexican cohort, higher alcohol consumption was associated with higher mortality.