In the detection of shunt and regurgitation in cardiac diseases or in the measurement of cardiac output by dye dilution method, an intuitive interpretation of the shape of obtained curves is unavoidable. Therefore, 'the results obtained are not always accurate. Especially, the curves recorded through ear piece after injection of the dye into the peripheral veins are sometimes equivocal and less accurate than those obtained by injection of dye in the large central vessels, although the former procedure is much simple and does not inflict much discomfort to the patients. Accordingly, it would be very advantageous for us to develop a means for deriving the useful information sufficient for clinical diagnosis from the curves obtained by the former method. In this respect, the writer has developed a method, in which the dye dilution curves recorded through ear piece is measured automatically and the presence or absence of shunt or regurgitation is determined by a computer. Calculation of cardiac output is, of course, included in the procedure. The dye dilution curves, which were used for the present study, were recorded by the Cam-bridge's Dye Dilution Curve Recorder using an ear piece. The employed dye was Coomassie blue, 40 mg of which was injected into the antecubital vein. In recording the dye dilution curve, generally to, ti, tp, tL, tr, td, cp, ct, cr and the terminal concentration (Et) have 'been measured hitherto. In writer's experience, however, the ascending and descending limbs also show outstanding characteristic in some of the diseases. Accordingly, the maximum slope of the ascending limb (maxS1) and its average slope (meanS1) during the period of 3 sec. (1.5 sec. before and 1.5 sec. after the point of the maximum slope), the maximum slope of the descending limb (maxS2) and its average slope (meanS2) were added to measuring items. When shunt or regurgitation exists, the calculation of td becomes inaccurate, so that the point (td, ) where meanS2 intersects the baseline was used in place of td, A ratio of Cp/Et was also added. Thus, a total of 19 items were used. The dye dilution curve was digitized by A-D converter at a sampling rate of 6 per second. The curves, which had been recorded on the sheets previously, were put in the tape recorder using a pencil follower, then delivered also to an A-D converter.