In a large proportion of femoral arteriograms, excellent filling of the veins of the thigh can be demonstrated. No controlled study appears to have been undertaken to establish whether or not this finding is normal. It has been suggested that “early” venous filling is due to arteriovenous anastomoses (1–3) and indicates one or all of the following: (a) arterial blockage; (b) inadequate anastomoses; (c) reduced arterial flow rate. Our impression, based upon a series of over 500 femoral arteriograms, is that no such relationship exists. The present study is an investigation of this so-called “early venous filling”: Does it bear any relationship to the above factors or is it possibly iatrogenic? The term, “early venous filling,” has been chosen in preference to arteriovenous shunting, since it has not been proved that shunts are responsible for early visualization of the venous circulation or that they are in fact pathological. A good deal of evidence suggests that arteriovenous anastomoses may be normal (4–6). Murphy and Margulis believed that a normal arteriogram without any sign of arteriovenous communications was virtually nonexistent and that the question as to whether or not they are pathological is merely a matter of degree. The arteriovenous anastomoses reported by several authors may well be the “clubbed” arteries or microaneurysms described by Margulis et al. as manifestations of occlusive peripheral vascular disease (8). There is also some doubt as to whether the term “early” is justifiable, since the actual time at which contrast medium may be expected to appear in the femoral veins after superficial femoral artery injection has not been satisfactorily established. Piulachs and Vidal-Barraquer ,(9) suggested that passage of the contrast agent to the veins is “very slow,” taking over thirteen seconds in normal subjects. These authors were apparently referring, however, to the time taken for the medium to appear in the veins of the calf, rather than those of the thigh. In addition, their injections were made with small quantities of medium (10 cc) injected very slowly (in four to five seconds). Their estimates, therefore, cannot be accepted as accurate for the normal femoral arteriogram. Within the three-year period covered by this study we have many times observed filling of the thigh veins where there has been no evidence of arterial disease. We have taken 25 of these cases as “normal” controls and have found on analysis that venous filling occurred at any time from four to twelve seconds after injection, with an average time of appearance of eight seconds. For the purpose of this article the term, “early venous filling,” is applied only to the clear demonstration of contrast medium in the veins of the thigh within two to six seconds after arterial injection (Fig. 1). Materials and Methods