Optimal IV-DSA performance depends equally on the digital system, the patient, and the radiologist. Through enhancement and subtraction, the digital system increases contrast sensitivity, thus compensating for the relatively low spatial resolution and for the loss in contrast (density) that results from dilution of the contrast medium by injecting it on the venous side. The degree of this dilution is governed by the patient's cardiac output and the size of the central blood volume. The lower the cardiac output and the larger the central blood volume, the less opacification (more dilution) and longer transit time (more likelihood for artifacts) will result. The role of the radiologist is to optimize the available conditions. He can prevent measures that decrease cardiac output (Valsalva maneuver) or he can take measures to decrease the degree of dilution by choosing optimal contrast administration methods, such as injecting in the right atrium at a high rate and thus allowing more latitude to decrease the total volume per injection and to increase the number of injections per examination. The radiologist also attempts to combat all the sources of noninformation or misinformation resulting from voluntary or involuntary patient motion, which degrades subtraction. By observing studies in real time, the radiologist may recognize motion during the injection, and by increasing the number of exposures, he may have a late mask to save the study. On immediate review of an injection, he may recognize the need to increase the volume per injection to obtain better opacification or, conversely, to reduce the volume if it is apparent that it could be done without compromise to the study and yet allow more injections to be performed, or he may recognize a finding that requires more than the usual routine views and may obtain a better one. IV-DSA can be performed with peripheral injections of contrast medium if the area of interest is limited and superior opacification is not essential, especially if a larger image intensifier (12- to 16-inch field) is available. However, whether peripheral injections are the least invasive, depends on whether one thinks that 200 to 240 ml of Renografin-76 for four views in an outpatient is invasive or not. The notion that IV-DSA can be performed by a nurse who places the IV needle or angiocatheter, a technician who shoots the pictures, and a computer that will do the rest, while the radiologist is peripherally involved and to hope for consistently good studies, is far from realistic.(ABSTRACT TRUNCATED AT 400 WORDS)