Background: Peripheral arterial disease in patients with type 2 diabetes mellitus is an important risk factor for vascular events. Recommendations about whether ankle-brachial index should be performed differ depending on the source; therefore, it is necessary to re-evaluate the most important risk factors associated with peripheral arterial disease and whether it is useful to perform ankle-brachial index in newly diagnosed and drug-naïve patients with diabetes, independent of age or peripheral arterial disease symptoms. Methods: A total of 711 subjects were divided into groups: group 1, 600 type 2 diabetes mellitus patients, symptomatic or not for peripheral arterial disease; group 2, 61 type 2 diabetes mellitus patients newly diagnosed and drug naïve; and group 3, 50 subjects without diabetes. Ankle-brachial index, medical records and physical examination were performed in all patients, accessing cardiovascular risk factors. Results: Analysing group 1 asymptomatic patient to peripheral arterial disease, we found abnormal ankle-brachial index in 49% (77/156) ⩾50 years and 42% (16/38) <50 years ( p = not significant). Considering drug-naïve patients, a peripheral arterial disease prevalence of 39% (24/61) was found; among these, 48% (13/27) were <50 years and 32% (11/34) were ⩾50 years ( p = not significant). A forward stepwise regression model was developed, with type 2 diabetes mellitus duration ( r2 = 0.12) and sedentary lifestyle ( r2 = 0.14) found as independent variable predictors of severity of peripheral arterial disease, related to ankle-brachial index. Conclusion: We suggest that, in type 2 diabetes mellitus, ankle-brachial index should be measured at diagnosis. In addition, sedentary lifestyle was strongly associated with presence and severity of peripheral arterial disease.