Methods: We analyse scanner data from a panel survey of demographically representative households (n = 885) collected over a one-year period (24 Jan 2010–22 Jan 2011) in the state of Victoria, Australia, which includes detailed records of each household's off-trade alcohol purchasing.

Findings: The heaviest consumers (3% of the sample) currently purchase 20% of the total litres of alcohol (LALs), are more likely to purchase cask wine and full strength beer, and pay significantly less on average per standard drink compared to the lightest consumers (A$1.31 [95% CI 1.20–1.41] compared to $2.21 [95% CI 2.10–2.31]). Applying a MUP of A$1 per standard drink has a greater effect on reducing the mean annual volume of alcohol purchased by the heaviest consumers of wine (15.78 LALs [95% CI 14.86–16.69]) and beer (1.85 LALs [95% CI 1.64–2.05]) compared to a uniform volumetric tax (9.56 LALs [95% CI 9.10–10.01] and 0.49 LALs [95% CI 0.46–0.41], respectively). A MUP results in smaller increases in the annual cost for the heaviest consumers of wine ($393.60 [95% CI 374.19–413.00]) and beer ($108.26 [95% CI 94.76–121.75]), compared to a uniform volumetric tax ($552.46 [95% CI 530.55–574.36] and $163.92 [95% CI 152.79–175.03], respectively). Both a MUP and uniform volumetric tax have little effect on changing the annual cost of wine and beer for light and moderate consumers, and likewise little effect upon their purchasing.

Conclusions: While both a MUP and a uniform volumetric tax have potential to reduce heavy consumption of wine and beer without adversely affecting light and moderate consumers, a MUP offers the potential to achieve greater reductions in heavy consumption at a lower overall annual cost to consumers.