There is emerging awareness of alcohol as a cause of the persisting health divide between east and west Germany. This study quantifies the burden of alcohol attributable mortality in the two parts of Germany in the 1990s, taking account of both adverse and beneficial effects of alcohol. We used an epidemiological approach that applies cause-specific alcohol attributable fractions derived from published relative risks and data on the distribution of alcohol consumption in east and west Germany in 1990/1992 and 1998 to mortality data for the two regions in 1992 and 1997, thus producing an estimate of the number of alcohol attributable deaths 'caused' or 'prevented'. Including the cardio-protective effect of alcohol, there were about 1.4% more deaths among men aged 20+ in 1992 in Germany than would have been expected in a non-drinking population, while there were 0.1% fewer deaths among women. By 1997, this had increased to 1.8% excess male deaths and 0.1% excess female deaths. In 1997, alcohol 'caused' 9.0% of all deaths in east German men compared with 5.6% in the west (women east: 2.5%; women west: 2.2%). At the same time, alcohol 'prevented' 5.2% deaths in east German men compared with 4.3% in the west, while there were 2.9% and 2.0% fewer deaths in women. This resulted in a net excess of deaths due to alcohol, except east German women, where 0.3% deaths were estimated to have been averted by alcohol. Although by 1997 net deaths 'caused' by alcohol had increased in the west and declined in the east, the burden of mortality due to alcohol among men remained highest in the east whereas in women the order had reversed.Mortality attributable to alcohol contributes considerably to overall mortality and to the east-west gap in Germany. This study points to the need for comprehensive policies on alcohol in Germany to close the persisting east-west health gap.