ObjectiveTo explore whether the inverse association between birth weight and mortality from circulatory diseases is mediated through blood pressure in men aged 50–75 years.DesignCohort study with retrospectively collected data on size at birth.Subjects and settingThe study included 1334 men born during 1920–1924, living in Uppsala, Sweden, who were examined at the ages of 50 and 60 years, and followedup to the end of 1995.Main outcome measuresMortality from circulatory diseases based on routine death registration.ResultsBirth weight showed a specific, inverse association with mortality from circulatory diseases: the rate ratio was 0.67 (95% confidence interval 0.50 to 0.89) per 1000 g increase in birth weight. This association was not appreciably affected by adjustment for sociodemographic characteristics or smoking, but was strengthened slightly by adjustment for body mass index at the ages of 50 and 60 years. Adjustment for systolic blood pressure at the age of 50 years only slightly reduced the strength of the inverse association between birth weight and mortality from ischaemic heart disease, and did not affect the inverse association between birth weight and mortality from stroke. Adjustments for systolic and diastolic blood pressure and hypertension treatment at the ages of 50 and 60 years did not reduce the strength of the association between birth weight and mortality from circulatory diseases at the age of 60–75 years.ConclusionsThe inverse association between birth weight and mortality from circulatory diseases in men aged 50–75 years is independent of adult sociodemographic characteristics, smoking and adult obesity and does not seem to be mediated through an increased blood pressure in those with low birth weight.