To observe the effects of anatomical distribution of left ventricular hypertrophy (LVH) and the differences in blood pressure on cardiac function, borderline hypertensives (group A) and mild or moderate sustained hypertensives (group B) were studied at rest and during dynamic exercise and their results were compared with those of normal controls (group C). LVH was estimated by M-mode echocardiography, and cardiac function was assessed by radionuclide (RI) first pass and multigate methods. In each subjects, 0.5 watt/kg and 1.0 watt/kg loads of ergometer exercise were administered for 16 min in the supine position. The main results were as follows: 1. In borderline hypertension, there was significant hypertrophy of the interventricular septum, and the left ventricular mass index was greater than that in normal controls. 2. The cardiac index in borderline hypertension did not differ from that of normal controls, but that of sustained hypertension was significantly reduced. 3. Significant positive correlations were observed between blood pressure and the left ventricular mass index at rest and during exercise. 4. No difference in left ventricular ejection fraction (LVEF) at rest was observed among the three groups, but that in sustained hypertension during exercise was markedly less than that of the other groups. LVEF at 1.0 watt/kg load and delta LVEF correlated inversely with the left ventricular mass index. 5. Left ventricular end-systolic volume determined by an RI count method decreased significantly during exercise in normals and borderline hypertension, while that in sustained hypertension was not reduced during exercise. Accordingly, cardiac function in the latter group was judged to be impaired. 6. In borderline hypertension, regional LVEF in the septal portion was slightly reduced at rest and during exercise. This reduction might have been caused by an imbalance of hypertrophy between the septal and free wall portions. In sustained hypertension, regional septal LVEF was maintained at rest, but was markedly less during exercise as compared with the other groups. Regional LVEF in the free wall portion was similar in degree to that of LVEF at rest and during exercise in the three groups. 7. Though the peak filling rate (PFR) in borderline hypertension did not differ from that in controls, that of sustained hypertension was significantly reduced at rest and during exercise. Significant positive correlations were obtained between PFR and LVEF at rest and during exercise. In summary, elevated blood pressure at rest and during exercise was thought to be the most important causative factor of LVH in hypertension.(ABSTRACT TRUNCATED AT 400 WORDS)