Current guidelines recommend obtaining an ankle–brachial index (ABI) to screen for peripheral artery disease (PAD) in subjects at risk. Previous work demonstrated that a combination of β₂-microglobulin, cystatin C, high-sensitivity C-reactive protein and glucose was associated with PAD. This study evaluated the ability of these biomarkers combined with clinical parameters to predict PAD in at-risk subjects. This study enrolled 1025 subjects from 99 primary care clinics who were smokers and/or diabetics ≥ 50 years or any individual ≥ 70 years. Consented subjects underwent a clinical assessment, fasting blood draw, and an ABI measurement with PAD defined as an ABI < 0.90 in either leg. The biomarkers and their interactions were evaluated using logistic regression and performance was evaluated at a cut point of the biomarker panel selected to maximize sensitivity while minimizing the false positive rate of the test. Of the 1025 subjects enrolled, 46 did not meet the ABI or other criteria for inclusion in the analysis. Among the evaluable subjects (n = 979), PAD was detected in 83 (8.5%). The model had a C-statistic of 0.73 (95% CI 0.67–0.79). There were 20 patients with PAD who were judged to be at low to moderate risk for cardiovascular events by clinical assessment; the model correctly identified 17 of these 20 patients. The model also performed well in subjects with no prior history of PAD. Thus, a biomarker panel may have a role for identifying PAD.