A national survey was performed in France from May to June, 1993. The aim of this study was to evaluate general practitioners' attitudes and behaviors when diagnosing and managing patients with lower extremity arterial disease (LEAD). One thousand general practitioners, randomly drawn from an exhaustive list, were contacted to participate in a telephone interview concerning the last patient with intermittent claudication seen in their practice. Four hundred seventy-six general practitioners participated. Risk factors noted for these 476 patients with intermittent claudication were in agreement with the literature: 86% were men aged 64 +/- 10 years (mean +/- SD) and 14% were women aged 73 +/- 8 years. Sixty-two percent had a pain-free walking distance of between 100 and 500 meters at diagnosis. Forty-five percent were former smokers and 37% currently smoked; 55% had hypertension, 14% diabetes, and 56% disturbances of lipid metabolism. A majority of them were hypercholesterolemic. The diagnosis of the disease was based primarily on a clinical assessment, confirmed for 33% by Doppler or echo Doppler. The mean duration of diagnosis was 4.4 +/- 4.1 years. Management of the disease was mainly by prescription of vasodilators (91%), antiplatelet agents (59%), and anticoagulants (8%). Use of Doppler or echo Doppler was recommended once a year. Infection was observed in 27% of patients. Thirty-eight percent had had a cardiac incident (angina pectoris or myocardial infarction) and 10% a cerebrovascular accident. They differed significantly from those with LEAD alone for the following parameters: age (68.5 +/- 9.2 vs. 63.2 +/- 10.3 years; p < 0.001); duration of LEAD (5.6 +/- 4.6 vs. 3.6 +/- 3.5 years; p < 0.001); hypertension (65% vs. 50%; p < 0.01); and current smoking (29% vs. 43%; p < 0.01). This survey confirmed the feasibility of telephone interviewing, on a large sample of general practitioners in France. The high level of association with other cardiac incidents was, for these patients, a much higher risk of mortality and morbidity than LEAD alone. It would be interesting to validate the associations observed with a prospective study of comorbidity.