Background Peripheral arterial disease (PAD) is an indicator of widespread atherosclerosis. However, most individuals with PAD, in spite of being at high cardiovascular risk, are asymptomatic. This fact, together with the limitations of the Doppler ankle-brachial index (ABI), contributes to PAD underdiagnose. The aim of this study was to compare oscillometric ABI and Doppler ABI to diagnose peripheral arterial disease, and also to examine the influence of oscillometric errors and calcified legs on the PAD diagnoses. Methods and Findings We measured the ankle-brachial indexes of 90 volunteers (n = 180 legs, age 70 ± 14 years, 43% diabetics) using both oscillometer OMRON-M3 and Doppler. For concordance analyses we used the Bland and Altman method, and also estimated the intraclass correlation coefficient. Receiver Operating Characteristic Curves were used to examine the diagnostic performance of both methods. The ABI means were 1.06 ± 0.14 and 1.04 ± 0.16 (p = 0.034) measured by oscillometer and Doppler ABIs respectively, with limits of agreement of ± 0.20 and intraclass correlation coefficient = 0.769. Oscillometer yielded 23 "error" measurements, and also overestimated the measurements in low ankle pressures. Using Doppler as gold standard, oscillometer performance for diagnosis of PAD showed an Area Under Curve = 0.944 (sensitivity: 66.7%, specificity: 96.8%). Moreover, when considered calcified legs and oscillometric "error" readings as arteriopathy equivalents, sensitivity rose to 78.2%, maintaining specificity in 96%. The best oscillometer cut-off point was 0.96 (sensitivity: 87%, specificity: 91%, positive likelihood ratio: 9.66 and negative likelihood ratio: 0.14). Conclusion Despite its limitations, oscillometric ABI could be a useful tool for the diagnosis of PAD, particularly when considering calcified legs and oscillometric "errors" readings as peripheral arterial disease equivalents.