Several stress tests have been used to predict the development of hypertension. No conclusion, however, has been reached on the effectiveness of these tests as a predictor of hypertension in later years. The present study examines the prognostic values of blood pressure response to cold and resting blood pressure for future hypertension, based on a follow-up study over 28 years of 824 individuals (mean age: 35.8 +/- 10.8 yr) whose resting blood pressure had been normal at baseline. A significant determinant of blood pressure response to cold stimulus was age in these normotensive subjects. A higher response was observed as age increased. There was also a significant seasonal variation in blood pressure response to cold, suggesting the need to standardize a time to perform the cold pressor test. Hypertension has developed in 343 individuals during the 28 years of follow-up, with a mean incidence rate of 24.6 per 10(3) person-years. Both systolic and diastolic responses were significant as a predictor of future hypertension after adjusting for attained age, resting blood pressure, and body mass index at baseline. However, a comparison between resting blood pressure and response to cold indicated that the cold pressor test is not as effective a predictor of hypertension as resting blood pressure. Radiation exposure was not significant either as a background risk or as a possible modifier of the relationship between the blood pressure response and the development of hypertension. The current results suggest that blood pressure response to cold supplements resting blood pressure for predicting hypertension.