Exposing a vein to altered hemodynamics by creating an arteriovenous (AV) shunt evokes considerable vessel formation that may be of therapeutic potential. However, it is unclear whether the introduction of oscillatory flow and/or flow increase is decisive. To distinguish between these mechanical stimuli we grafted a femoral vein into the arterial flow pathway of the contralateral limb in rats creating an arterioarterial (AA) loop (n = 7). Alternatively, we connected the femoral artery and vein using the vein graft, whereby we created an AV-loop (n = 27). Vessel loops were embedded in a fibrin filled chamber and blood flow was measured by means of flow probes immediately after surgery (day 0) and 15 days after loop creation. On day 15, animals were sacrificed and angiogenesis was evaluated using μCT and histological analysis. Mean flow increased from 0.5 to 2.4 mL/min and was elevated throughout the cardiac cycle at day 0 in AV-loops whereas, as expected, it remained unchanged in AA-loops. Flow in AV-loops decreased with time, and was at day 15 not different from untreated femoral vessels or AA-loop grafts. Pulsatile flow oscillations were similar in AV-and AA-loops at day 0. The flow amplitude amounted to ~1.3 mL/min which was comparable to values in untreated arteries. Flow amplitude remained constant in AA-loops, whereas it decreased in AV-loops (day 15: 0.4 mL/min). A large number of newly formed vessels were present in AV-loops at day 15 arising from the grafted vein. In marked contrast, angiogenesis originating from the grafted vein was absent in AA-loops. We conclude that exposure to substantially increased flow is required to initiate angiogenesis in grafted veins, whereas selective enhancement of pulsatile flow is unable to do so. This suggests that indeed flow and most likely wall shear stress is decisive to initiate formation of vessels in this hemodynamically driven angiogenesis model.