Age standardised death rates (European standard population) from lung cancer in the Russian Federation, have been rising since at least 1965, levelled out in the late 1980s and have subsequently decreased. The reasons for this decline are not apparent. This study seeks to identify the reasons for the decline in mortality from lung cancer in the Russian Federation in the 1990s. Changes in age-specific mortality from lung cancer in the Russian Federation between 1990 are described and age-cohort analysis, based on age-specific death rates for lung cancer is undertaken for the period 1965 to 1995. As other work has shown that any recent deterioration in coding of cause of death has been confined largely to the elderly, this suggests that the trend is not a coding artefact. Age-period-cohort analysis demonstrates the existence of a marked birth cohort effect, with two major peaks corresponding to those born around 1926 and 1938. These groups would have reached their early teens during the second world war and the period immediately after the death of Stalin, respectively. The present downward trend in death rates from lung cancer in the Russian Federation is partly due to a cohort effect and it is expected that this will soon reverse, with a second peak occurring in about 2003.