Background and aims: When measuring inequalities in health, public health and addiction research has tended to focus on differences in average life‐span between socio‐economic groups. This does not account for the extent to which age of death varies between individuals within socio‐economic groups or whether this variation differs between groups. This study assesses (1) socio‐economic inequalities in both average life‐span and variation in age at death, (2) the extent to which these inequalities can be attributed to alcohol‐specific causes (i.e. those attributable only to alcohol) and (3) how this contribution has changed over time.

Design: Cause‐deleted life table analysis of national mortality records.

Setting and cases: England and Wales, 2001–16. All‐cause and alcohol‐specific deaths for all adults aged 18+, stratified by sex, age and quintiles of the index of multiple deprivation (IMD).

Measurements: Life expectancy at age 18 years and standard deviation in age at death within IMD quintiles and the contribution of alcohol to overall differences in both measures between the highest and lowest IMD quintiles by comparing observed and cause‐deleted inequality ‘gaps’.

Findings: In 2016, alcohol‐specific causes reduced life expectancy for men and women by 0.26 and 0.14 years, respectively, and increased the standard deviation in age at death. These causes also increased the inequality gap in life expectancy by 0.33 years for men and 0.17 years for women, and variation in age at death by 0.14 years and 0.13 years, respectively. For both measures, the contribution of alcohol to mortality inequalities rose after 2001 and subsequently fell back. For women, alcohol accounted for 3.6% of inequality in age at death and 6.0% of life‐span uncertainty, suggesting that using only the former may underestimate alcohol‐induced inequalities. There was no comparable difference for men.

Conclusions: Deaths from alcohol‐specific causes increase inequalities in both life expectancy and variation in age of death between socio‐economic groups. Using both measures can provide a fuller picture of overall inequalities in health.