Longitudinal data from different populations have shown different degrees of tracking of blood pressure (BP). To examine BP tracking in Korean school children, 219 of 430 children (100 males, 119 females) who were 6 years old in 1986 in Kangwha County, Korea have been examined annually up to 1997 in the Kangwha Study. BP was measured twice with standard mercury sphygmomanometers and the average of the two measurements was used for the level of BP. Diastolic BP were measured at the fourth Korotkoff sound. Tracking was examined using a time-lag correlation analysis and McMahan's tracking index tau, which indicates the proportion of variation attributable to tracking apart from the natural growth component. As well the jackknife method was used to obtain the confidence interval of tau. Correlation coefficients between systolic BP from age 6 to 17 ranged from 0.39-0.54 for males and 0.44-0.57 for females. Taus for systolic BP were 0.875 (95% CI: 0.803-0.947) and 0.900 (95% CI: 0.809-0.991) in males and females, respectively. Correlation coefficients between diastolic BP from age 6 to 17 ranged from 0.28-0.47 for males and 0.14-0.47 for females. Taus for diastolic BP were 0.983 (95% CI: 0.897-1.000) and 0.800 (95% CI: 0.717-0.883) in males and females, respectively. These findings showed strong evidence for BP tracking in Korean school children from childhood to late adolescence.