The aim was to examine the risk profiles and prognosis of treated and untreated hypertensive subjects and examine to what degree confounding by indication was present in a population-based cohort study with up to 30-year follow-up. The study population consisted of 9328 men and 10 062 women, aged 33-87 years at the time of attendance from 1967 to 1996. The main outcome measures were myocardial infarction (MI), cardiovascular disease (CVD) mortality and all-cause mortality. Comparing the risk profiles between treated and untreated subjects entering the study showed significantly higher values for some risk factors for treated subjects. During the first 10 years, hypertensive men without treatment, compared with those treated, had a significantly lower risk of suffering MI, CVD and all-cause mortality, hazard ratio (HR) 0.72 (95% CI; 0.57, 0.90), 0.75 (95% CI; 0.59, 0.95) and 0.81 (95% CI; 0.61, 0.98), respectively. No significant differences in outcome were seen during the following 20 years. In identically defined groups of women, no significant differences in mortality were seen between groups. Subgroup analysis, at two stages of the study 5 years apart, revealed that some cardiovascular risk factors had a higher prevalence in hypertensive men who were treated at the later stage, compared with those who remained untreated (P=0.004). In conclusion, hypertensive treated men had a worse prognosis during the first 10 years of follow-up than untreated ones, which is most likely due to worse baseline risk profile. Hypertensive men that were treated at a later stage had a worse risk profile than those not treated at a later stage.