The prevalence of signs of heart involvement was studied non-invasively in a group of untreated hypertensives (n=35) and a reference group (n=73), all derived from a random population sample of 50-year-old men. Signs of left ventricular hypertrophy were studied by means of orthogonal electrocardiography and conventional electrocardiography. Signs of decreased distensibility of the left ventricle were studied by apex cardiography and registration of atrial sounds. Left ventricular hypertrophy among hypertensives was significantly more common according to orthogonal electrocardiography (33%) than according to conventional electrocardiography (9%), indicating that the former may be a better method for detection of left ventricular hypertrophy than the latter. In the hypertension group the amplitude of the R wave in lead X on orthogonal electrocardiography was positively correlated to casual diastolic blood pressure (r=0-40) and to diastolic blood pressure after one hour's rest (r=0-65). The degree of pressure load leading to left ventricular hypertrophy seems to be better reflected by resting than by casual blood pressure. There was no hypertensive subject with both signs of left ventricular hypertrophy on orthogonal electrocardiography and either an a/H ratio over 15 per cent or an abnormal atrial sound, indicating two different forms of cardiac involvement as the result of hypertension. Casual blood pressures became normal during rest in hypertensives with a/H ratio over 15 per cent on apex cardiography or abnormal atrial sound, not in hypertensives with signs of left ventricular hypertrophy on orthogonal electrocardiography.