Objective: Population blood pressure (BP) levels from a longitudinal study were analysed for trends during a period of 15 years. Trends from unadjusted data are reported as well as trends adjusted for major cardiovascular (CV) risk factors and use of antihypertensive therapy, thus allowing assessment of independent BP trends. Design: The Copenhagen City Heart Study is a longitudinal epidemiological study of CV risk in a random population sample of both genders aged 20 and above. Three cross-sectional population surveys were performed: 1976-78 (n = 14000), 1981-83 (n = 12675) and 1991-94 (n = 9661). Methods: BP was measured by a London School of Hygiene Sphygmomanometer. Weight and height were measured and body mass index (BMI) calculated. Non-fasting plasma cholesterol was determined. A questionnaire concerning smoking status and diabetes was completed. Measurement methods were strictly standardized and unchanged in the three cross-sectional surveys. Results: Unadjusted systolic BP (SBP) levels decreased during 15 years of follow-up, and unadjusted diastolic BP (DBP) levels increased. An investigation of the effect of major CV risk factors, both singly and jointly on BP levels, revealed a pattern of correlations contributing to BP variability. Adjustments for BMI, cholesterol, diabetes, use of antihypertensive therapy and smoking status were made in the final analyses of BP trend. The adjusted trend model demonstrated that SBP levels remained lower than SBP levels in the first survey. DBP levels increased slightly. Conclusions: The results demonstrate a decrease in population SBP. The decrease is independent of major CV risk factors. Possible contributing factors are discussed.