Cell-based therapy has been proposed as a novel strategy for patients with severe peripheral arterial disease by stimulating vascular growth. In clinical studies of this therapy, the emphasis has been on demonstrating recovery of clinical parameters, rather than on evaluation of blood flow recovery. Angiography is still the gold standard for the assessment of lower leg arteries. Therefore, we studied the usefulness of angiography in the evaluation of cell-based therapy. Sixteen patients with critical leg ischemia (ischemic rest pain or ulcers), or persistent (>12 months) profound disabling claudication were unilaterally treated with autologous bone marrow-derived mononuclear cells. Pre- and 6 months post-treatment digital subtraction angiographies (DSA) were assessed and compared in a blinded fashion twice by a panel of seven vascular surgeons and interventional radiologists. Inter- and intraobserver variability on qualitative (poor/moderate/rich) and semi-quantitative (increase/no difference/decrease) assessment of collateral circulation were evaluated. Agreement was expressed inter- and intraclass correlation coefficients (CC). Inter- and intraobserver agreement was moderate for the qualitative grading of collateral extent (CC = 0.46 and 0.60, respectively). Agreement was moderate (inter-CC = 0.60) to good (intra-CC = 0.73) for comparing pre- and post-treatment DSA. Clinical response was based on limb salvage, pain-free walking distance, ankle-brachial pressure index and pain scores. No difference was observed in the extent of collateral circulation between pre- and post treatment DSA after separate analysis of clinical responding and non-responding patients (P = 0.92). DSA is not a suited modality for the evaluation of therapeutic angiogenesis.