The relation between the apex cardiogram and changes in left ventricular pressure measured by micromanometer, and dimension measured by echocardiography, was studied in 12 normal subjects and 64 patients with heart disease. In 12 patients, the apex cardiogram was delayed with respect to simultaneous left ventricular pressure by 17±18 ms during the upstroke and 28±16 ms during the downstroke. In the normal subjects, changes in left ventricular dimension during the upstroke and downstroke of the apex cardiogram were small, amounting to 6±5 and 21 ± 7 per cent total excursion, respectively. In 10 patients with mitral regurgitation, there was significant inward wall movement during the upstroke and in 10 patients with aortic regurgitation, significant outward movement during the downstroke, both reflecting valvular regurgitation. In 20 patients with ischaemic heart disease and segmental abnormalities on left ventricular angiography, apex cardiogramecho dimension relations were abnormal in all, because of inward or outward wall movement during the upstroke, increased outward movement before the 'O' point, or abnormal inward movement during the downstroke. These disturbances were displayed by constructing apex cardiogram-echo dimension loops, which appear to be a sensitive means of detecting incoordinate left ventricular contraction, analogous to those between pressure and dimension.A limitation in the use of M-mode echocardiography in studying left ventricular function is that wall movement is recorded from only a localized region of cavity, so that it is not possible to tell whether a change in dimension results from filling or ejection on the one hand, or from an isovolumic shape change on the other. The value of the technique would, therefore, be increased if information were simultaneously available allowing this distinction to be made, and incoordinate contraction and relaxation detected and quantified. We have previously shown that these abnormalities may lead to characteristic alterations in the time relations between left ventricular dimensions measured by echocardiography and left ventricular pressure , particularly when a pressure-dimension loop is constructed. In the present study, we have explored the possibility of substituting the apex cardiogram for the pressure pulse in order to obtain the same information noninvasively. We have, therefore, investigated the time relations between the various phases of the