OBJECTIVE : Patients with infrainguinal peripheral arterial disease often undergo multiple revascularisation procedures. Although many centres have adopted an endovascular first approach, some are reluctant to do so for fear of compromising the outcomes of any subsequent bypasses. All studies that compared the outcomes of primary infrainguinal bypass with bypass after failed endovascular intervention were analysed. METHODS : A systematic review was conducted of MEDLINE, EMBASE, and CENTRAL databases for studies comparing outcomes of primary infrainguinal bypass with bypass after failed endovascular intervention for peripheral arterial disease. Abstracts and full text studies were screened independently by two reviewers with data abstraction done in duplicate. Dichotomous outcome measures were reported using the OR and 95% CI, and pooled using random effects models. RESULTS : Abstracts were screened (2,528), with 50 selected for full text review. Of these, 15 studies involving 11,886 patients met the inclusion criteria. Pooling the results of studies comparing primary bypass with bypass after failed endovascular intervention showed no significant difference in 30 day mortality (OR 1.00; 95% CI 0.65-1.54), or 30 day amputation rates (OR 1.26; 95% CI 0.95-1.65). Interestingly, one year amputation free survival was higher in the patients who had primary bypass (OR 1.30; 95% CI 1.10-1.52) compared with patients who had bypass after failed endovascular therapy. There was also worse one year primary patency (OR 1.65; 95% CI 1.04-2.62) for patients with prior failed endovascular intervention. The review demonstrated a trend towards higher rates of early graft occlusion (OR 4.54; 95% CI 0.97-21.28). CONCLUSIONS : Meta-analysis of the existing literature comparing primary bypass with bypass following failed endovascular intervention shows worse one year amputation free survival and worse primary patency in those patients who undergo bypass after failed endovascular intervention. There is also a trend towards higher rates of early graft occlusion, although these results were not statistically significant. These conclusions are limited by observational study design, inconsistent patient selection, and significant heterogeneity between studies.