STUDY OBJECTIVE: To examine the impact of a systematic risk factor screening and counselling carried out by family physicians and family nurses within the larger framework of a community intervention programme for the prevention of cardiovascular disease (CVD). DESIGN: Quasi-experimental study comparing trends in an intervention area with those in a reference area. SETTING: A Northern Sweden municipality (5500 inhabitants) constituted the intervention area while the Northern Sweden region (510,000 inhabitants) served as the reference area. PARTICIPANTS: All 30, 40, 50, and 60 year old inhabitants were invited each year from 1985 to 1992. Among 2046 eligible 1893 participated (92.5%), which formed eight independent cross sections. One cross section, 1986, was re-surveyed forming a panel. MAIN RESULTS: In the cross sections, mean total cholesterol was reduced from 7.09 to 6.27 mmol/l for men (p < 0.001) and from 7.13 to 5.89 mmol/l for women (p < 0.001) and mean systolic blood pressure from 132.2 to 123.7 mm Hg for men (p < 0.05) and from 129.2 to 122.0 mm Hg for women (p < 0.001) during the eight years. Body mass index (BMI) increased from 25.6 to 26.2 for men (p < 0.05) and from 25.0 to 25.5 for women (NS). A corresponding reduction in cholesterol and blood pressure (for women) occurred in the panel, while BMI was unchanged. The risk for CVD, using the Framingham equation, was estimated to be reduced overall by 19% (p = 0.0021) when comparing early cross sections (1985/86) with the later cross sections (1990/91). CONCLUSIONS: It was concluded that a long term community based CVD prevention programme that combines population and individual strategies can substantially promote a health shift in CVD risk in a high risk rural population. The individual attention and evaluation provided by the health provider survey seem to accelerate, but not increase the amount of, risk reduction.