OBJECTIVE To assess the utility of a novel rapid urinary cotinine assay to detect and quantify the level of smoking in patients with peripheral arterial disease. METHODS This was a cross-sectional study in a vascular surgical outpatient department of a large teaching hospital. Participants were 100 consecutive subjects presenting to a hospital outpatient clinic with a diagnosis of intermittent claudication confirmed by a positive Edinburgh claudication questionnaire and an ankle-brachial pressure index of less than 0.9. Main outcome measures were patient-offered smoking history, exhaled breath carbon monoxide levels, urinary cotinine levels as measured by a novel rapid assay, and laboratory-measured creatinine-adjusted urinary cotinine levels. Results Fifty-five subjects declared that they were current smokers, 40 declared that they were ex-smokers, and 5 declared that they were never-smokers. Of the 40 ex-smokers, 6 subjects (15%) had urinary cotinine levels greater than 500 ng/mL (regular smokers), and a further 2 (5%) had urinary cotinine levels between 100 and 500 ng/mL (light, irregular, or passive smokers). The rapid urinary cotinine assay had a sensitivity and specificity of 100% and 98%, respectively, in its ability to detect active smoking, and the degree of smoking correlated well with laboratory creatinine-corrected urinary cotinine levels (Spearman coefficient, 0.805; P < .001). By contrast, exhaled carbon monoxide had a sensitivity and specificity of 95% and 89%, respectively, and although it correlated well with urinary cotinine (Spearman coefficient, 0.782; P < .001), it was found on linear regression analysis to be unreliable in differentiating light smokers from nonsmokers. CONCLUSIONS Patient-offered smoking history is unreliable because there is no correlation between the patient-reported number of cigarettes smoked per day and urinary cotinine levels. The novel rapid assay for urinary cotinine described here is superior to exhaled carbon monoxide measurement in detecting the level of smoking exposure among patients with intermittent claudication, and its results correlate well with laboratory-measured cotinine.