OBJECTIVE: To assess whether guidelines for the management of suspected PE, recommending that ventilation/perfusion (V/Q) scintigraphy should be followed by other imaging in case of non-diagnostic result, and interpreted along with the clinical probability, are applied in daily practice. DESIGN: Two-year audit study. SETTING: Tertiary hospital in France. PARTICIPANTS: All patients referred to the nuclear medicine department for a suspected pulmonary embolism (PE), with a low clinical probability, a positive D-Dimer test, and a low V/Q scintigraphy. INTERVENTION: We reviewed medical records to collect data on further diagnostic strategy and therapeutic management. MAIN OUTCOME MEASURE: Thromboembolic risk during a three-month follow up in patients who did not undergo anticoagulation therapy on the basis of a negative diagnostic work up. RESULTS: Of the 456 selected patients, PE was excluded on the basis of a low pretest probability and a low V/Q scintigraphy probability without further testing in 184 (group 1). In the other 272 patients (group 2), 4 venous thromboembolism (VTE) events were diagnosed by leg vein ultrasonography and/or computed tomography pulmonary angiography. In patients who did not receive anticoagulant treatment during follow up, one patient from group 1 (0.6%, 95% confidence interval 0.1 to 3.3) and two patients from group 2 (0.9%, 95% CI 0.2 to 3.2) had an acute venous thromboembolic event. CONCLUSION: Different attitudes are adopted by physicians in patients with a low clinical probability and a low V/Q scintigraphy probability, without impact on safety, as assessed by a non-significant difference in the three-month thromboembolic risks between these two groups.