Aims. To assess changes in alcohol consumption attributable to the Great Hanshin Earthquake to test the assumption that alcohol consumption increases after natural disasters. Methods. Quarterly alcohol sales figures were compared for three periods: before, immediately after and subsequent to the Great Hanshin Earthquake in three areas of the Hyogo prefecture: the severely affected area, the moderately affected area and the unaffected area. Possible confounding by population movement, damage to retail outlets and normal variation in sales, was assessed. Findings. The quantity of alcoholic beverages consumed in the heavily damaged areas as well as throughout the prefecture decreased from the 1994 pre-disaster level, both immediately after the Great Hanshin Earthquake (January-March 1995) and 2 years after the disaster. This finding remained once possible confounding factors were taken into consideration. Conclusions. More attention should be placed on drinking in the cultural context where a disaster occurs. It may sometimes deter, rather than encourage, drinking among the affected population.