Although many studies have shown that smoking is associated with an increased risk of death from ischaemic heart disease (IHD), and that the increase appears to vary with age and amount smoked, there has been little formal specification or estimation of the relationship. In this paper two alternative models are tested, using data for different ages and levels of smoking from four major studies in three countries. One model explains 80% of the variation in mortality in terms of a positive linear function of the number of cigerettes smoked, the parameters of which decrease with age. We estimate that every cigerette smoked per day increases the risk of dying from IHD by as much as 35% at ages 35 to 44, reducing to 2% at ages 65 to 74. The risk attributable to smoking may account for more than 80% of IHD deaths of men aged 35 to 44, and 27% of those of men aged 45 to 64. Although the relative risk is highest for younger age groups, the absolute risk of death from IHD that is attributable to smoking increases with age. The evidence suggests that both are increasing with time.