Our previous studies have shown that when a flap is delayed, there is dilation of existing vessels within the flap not ingrowth of new vessels. The maximal anatomic effect on the arterial tree occurs at the level of the reduced-caliber "choke" anastomotic vessels that link adjacent vascular territories. To further investigate the sequence of anatomic changes that occurs during the delay phenomenon, a large series of 200 rabbits and 17 dogs underwent a flap delay procedure in either skin or muscle and the tissues were examined at postoperative periods between 1 hour and 1 year by using well-established fluorescein, angiographic, light microscopic, immunohistochemical, and electron microscopic techniques. These data in the rabbit skin consistently demonstrated an initial period of vasoconstriction that resolved within 3 hours postoperatively and was followed by an active and progressive dilation of choke vessels that was most dramatic between 48 and 72 hours. In vivo intravenous fluorescein dye testing revealed an interesting parallel in that there was a temporary barrier to the flow of fluorescein that occurred at the level of the choke vessels immediately after the flap was raised and that this temporary barrier-continued to impede the flow toward the flap tip in rabbits where flaps had been delayed for periods up to 72 hours. Thereafter, there was no obstruction to the flow of fluorescein along the flap. During this early delay period of 3 days, light microscopy revealed a decrease in vessel wall thickness associated with an increase in lumen diameter. Over the next 4 days, the luminal diameter continued to dilate to a lesser extent and the vessel wall thickened. Immunohistochemical analysis showed increased cell division, maximal between 24 and 72 hours, in all layers of the choke vessel wall. During this same postoperative interval, transmission electron microscopy revealed phenotypic changes in smooth muscle cells from contractile to synthetic cells. Hypertrophy of the smooth muscle cells was also observed. The vascular endothelium, which initially showed evidence of denudation, was restored to a healthy intact appearance within the first week after delay. When followed for longer periods, long-term studies of the delayed flap of up to 1 year demonstrated dramatically a permanent dilation of the choke vessel lumen and a thickening of the choke vessel wall. In canine studies, one rectus abdominis muscle was delayed by ligating the deep inferior epigastric artery. The time sequence of choke vessel dilation, studied by sequential angiograms in vivo, was comparable to that of the rabbit skin model. To ascertain the permanence and irreversibility of this dilation, the normal circulation of the delayed rectus abdominis muscle was restored by reanastomosing the deep inferior epigastric artery. Even after a recovery period of up to 3 months, the choke vessels remained dilated and tortuous instead of reverting to their original narrow diameters. From this work, it is suggested that the choke vessel dilation seen in the delay period is a permanent and irreversible event. It is an active process associated with both an increase (hyperplasia) and an enlargement (hypertrophy) of the cells in all layers of the choke artery wall and a resultant increase in caliber of these vessels. The time sequence for delay appears to be similar in different species and in different tissues, suggesting the possibility of a universal process for delay.