PURPOSE: The purpose of the study is to examine the association between longitudinal change in body mass index (BMI) and change in coronary heart disease (CHD) risk score by using the Framingham risk score equation. METHODS: A general adult population in the municipality of Tromsø, Norway, was invited to four consecutive examinations in 1979 to 1980, 1986 to 1987, 1994 to 1995, and 2001. A total of 10,214 men and women aged 20 to 61 years at baseline attended at least three times. Associations were examined by using fixed-effects regression methods for longitudinal data. RESULTS: We observed a significant association between BMI change and risk score change in all baseline age groups. The association was significantly strengthened by age in women, but not men. A BMI increase of 3 kg/m(2) in subjects aged 40 to 49 years was associated with risk score increases of 0.45 points (95% confidence interval [CI], 0.29-0.62) in men and 0.66 points (95% CI, 0.52-0.80) in women. CONCLUSIONS: The well-known increase in body weight is associated with adverse CHD risk in both men and women.