Although the angiogenic effect of vascular endothelial growth factor (VEGF) is widely recognized, a central question concerns whether the vessels formed on its overexpression effectively increase tissue perfusion in vivo. To explore this issue, here we exploit AAV vectors to obtain the prolonged expression of VEGF and angiopoietin-1 (Ang1) in rat skeletal muscle. Over a period of 6 months, muscle blood flow (MBF) and vascular permeability were measured by positron emission tomography and single-photon emission computed tomography, respectively. All measurements were performed under resting conditions and after electrically induced muscle exercise. Despite the potent angiogenic effect of VEGF, documented by vessel counting and intravascular volume assessment, the expression of this factor did not improve resting MBF, and it even decreased perfusion after exercise. This deleterious effect was related to the formation of leaky vascular lacunae, which accounted for the occurrence of arteriovenous shunts that excluded the downstream microcirculation. These effects were significantly counteracted by the coinjection of VEGF and Ang1, which determined a marked increase in resting MBF and, most notably, a significant improvement after exercise that persisted over time. Taken together, these results challenge the effectiveness of VEGF as a sole factor to induce angiogenesis and suggest the use of factor combinations to achieve competent vessel formation.