We analysed the association of body mass index (BMI) with blood pressure (BP) at baseline, whether BMI predicts the incidence of anti-hypertensive drug treatment during a 12-year follow-up and whether this risk is independent of the original BP level and, finally, how diabetes associates with the incidence of anti-hypertensive drug treatment. The study population comprised 15,438 men and women in eastern Finland aged between 30 and 59 years who were not using anti-hypertensive drug treatment during baseline surveys in 1972 and 1977. At baseline BP increased linearly by increasing BMI. The proportion of hypertensive subjects, defined as either DBP > or = 95 mm Hg or SBP > or = 160 mm Hg, was 18% among the leanest men, BMI < 20 kg/m2, but 61% among the most obese, BMI > or = 30 kg/m2. Among women these proportions were 11% and 54%, respectively. Among the normotensive subjects at baseline, the BMI associated risk ratio of the incidence of anti-hypertensive drug treatment, adjusted for age and study year, was 1.14 (per kg/m2; P < 0.001) in men and 1.11 (P < 0.001) in women. After a further adjustment for DBP and SBP at baseline, risk ratios were 1.13 (P < 0.001) and 1.07 (P < 0.001), respectively. Diabetes associated with the risk of anti-hypertensive drug treatment independently from BMI, DBP and SBP. Because BMI correlates with BP cross-sectionally, and it also predicts the future increase in BP independently from the baseline BP, excess weight is undoubtedly one of the most important risk factors for hypertension. Weight control is the most natural primary intervention method in the inter-relation of obesity, hypertension and diabetes and in the prevention of subsequent cardiovascular diseases.