OBJECTIVE--To assess the association between sodium and potassium intake and the rise in blood pressure in childhood. DESIGN--Longitudinal study of a cohort of children with annual measurements during an average follow up period of seven years. SETTING--Epidemiological survey of the population of a suburban town in western Netherlands. SUBJECTS--Cohort of 233 children aged 5-17 drawn at random from participants in the population survey. MAIN OUTCOME MEASURES--At least six annual timed overnight urine samples were obtained. The mean 24 hour sodium and potassium excretion during the follow up period was estimated for each participant and the sodium to potassium ratio calculated. Individual slopes of blood pressure over time were calculated by linear regression analysis. RESULTS--No significant association was observed between sodium excretion and the change in blood pressure over time. The mean systolic blood pressure slopes, however, were lower when potassium intake was higher (coefficient of linear regression -0.045 mm Hg/year/mmol; 95% confidence interval -0.069 to -0.020), and the change in systolic pressure was greater when the urinary sodium to potassium ratio was higher (0.356 mm Hg/year/unit; 95% confidence interval 0.069 to 0.642). In relation to potassium this was interpreted as a rise in blood pressure that was on average 1.0 mm Hg (95% confidence interval -1.65 to -0.35) lower in children in the upper part of the distribution of intake compared with those in the lower part. The mean yearly rise in systolic blood pressure for the group as a whole was 1.95 mm Hg. Urinary electrolyte excretion was not associated with diastolic blood pressure. CONCLUSION--Dietary potassium and the dietary sodium to potassium ratio are related to the rise in blood pressure in childhood and may be important in the early pathogenesis of primary hypertension.