BACKGROUND: The longitudinal relationship between coffee use and hypertension is not well known. Aim. We did a prospective study to investigate if there is a temporal relationship between coffee consumption and development of sustained hypertension. METHOD: We assessed 1107 white subjects with elevated blood pressure who were followed up for 6.4 years. Coffee intake and other life-style factors were ascertained from regularly administered questionnaires. Incident physician-diagnosed hypertension was the outcome measure. RESULTS: During the follow-up, 561 subjects developed sustained hypertension, whereas 546 subjects did not meet the criteria for treatment. Coffee drinkers developed sustained hypertension more frequently than abstainers (53.1% versus 43.9%, P = 0.007). The incidence of hypertension did not differ between moderate and heavy coffee drinkers. Kaplan-Meier analysis confirmed that sustained hypertension was developed more frequently by coffee drinkers compared with nondrinkers (P<0.001). The adjusted relative risk of hypertension was greater in both categories of coffee drinking than in abstainers (hazard ratio, 95% confidence limit (CL) = 1.24, 1.06-1.44). The risk of hypertension associated with coffee drinking increased gradually with increasing level of alcohol use (adjusted P for interaction = 0.005). CONCLUSIONS: In subjects screened for stage 1 hypertension a nonlinear association was found between coffee consumption and development of sustained hypertension.