The reliability of a modified videodensitometric and photodensitometric sampling technique for measuring phasic flow rates in the coronary artery system was examined. Electromagnetic flow measurements were performed in a circulatory model with continuous and pulsatile flow and intraoperatively in aortocoronary bypass grafts; cineangiograms were made simultaneously. Based on the front velocities of injected boluses of contrast medium, the densitometric measurement overestimated the electromagnetically measured flow systematically by about 20%. Systolic and diastolic flow rates in aortocoronary bypass grafts and coronary arteries determined from biplane cineangiograms in 34 patients generally revealed the typical pulsatile flow pattern familiar from electromagnetic and ultrasonic flow measurements. Flow velocities in unstenosed coronary arteries were nearly identical before and after branchings of the vessels, whereas the corresponding flow rates were higher in proximal than in distal segments. The identical flow velocities in different branches of the same vessel and the low variability of this parameter in different patients may be a suitable index of the effect of stenoses on coronary arterial blood flow. Circulation 68, No. 2, 337-347, 1983. THE SEVERITY of coronary artery disease is currently estimated by subjective evaluation of morphologic vessel abnormalities visualized by coronary angiography. Several methods have been used in the past to provide a quantitatiye means of measuring coronary blood flow in man, including indicator-dilution and radioisotope techniques as well as Doppler ultrasonic flow catheter measurements and densitometric evaluations of coronary angiograms. 1-8 Generally x-ray densitometry is based on the determination of the mean transit time of contrast medium. Mean transit time is defined as the difference between the mean appearance times of the contrast medium measured from the "densograms" (time function of x-ray density) at a proximal site and a distal site over the vessel.95In an attempt to avoid several methodologic difficulties inherent in this technique, we determined transit times from the fronts of the densograms (appearance From time) instead of the mean appearance time. The flow rates determined from the front velocities were compared with those flow values measured electromagnetically during the interval of the passage of the contrast medium. The measurements were performed in a model circulation and during coronary revascularization. From these examinations, the systematic deviation between electromagnetic and densitometric flow measurements was determined. In addition, the influence of the injected contrast medium on coronary artery flow was evaluated. The aim of this study was to establish the methodologic requirements for densitometric measurements of systolic and diastolic flow in the coronary artery system.