Abnormal left ventricular (LV) diastolic relaxation is an early sign of hypertensive heart disease. Whether LV diastolic dysfunction is caused directly by raised blood pressure, or by structural changes related to LV hypertrophy remains controversial. We examined 115 hypertensive patients with LV hypertrophy, and two ageand gender-matched groups (38 hypertensive patients without LV hypertrophy and 38 normotensive subjects) by echocardiography to assess determinants of LV diastolic function, and the relation between diastolic function and LV geometric pattern. Diastolic function was evaluated by the E/A-ratio, E wave deceleration time (E-dec), isovolumic relaxation time (IVRT), and the atrioventricular plane displacement method (AV-LA/AVmean). A multivariate analysis (including gender, age and body mass index) shows diastolic function to be inversely related to blood pressure, LV wall thickness and LV mass, but not to LV end diastolic diameter. The E/A-ratio generally showed the strongest relations. Only the E/A-ratio and AV-LA/AV-mean were related to heart rate. By stepwise regression analysis, age was the strongest determinant for the E/A-ratio, E-dec and AV-LA/AV-mean, followed by systolic blood pressure, heart rate and LV wall thickness. For IVRT, however, LV wall thickness appeared strongest, followed by systolic blood pressure and age. In conclusion, blood pressure and LV wall thickness both have independent influence on LV diastolic function. Age and blood pressure are the most important factors to determine the E/A-ratio and Edec, whereas LV geometry and blood pressure are most important when IVRT is used. AV-LA/AV-mean may not be useful in hypertensive LV hypertrophy.