Considerable progress has been made in pulmonary embolism diagnosis during the last ten years. New diagnostic instruments such as plasma D-dimer measurement and lower venous compression ultrasonography have been developed. Clinical evaluation of the likelihood of pulmonary embolism has been rehabilitated and proven accurate and valid. The interpretation of lung scan results has become more simple and clear to clinicians. Finally, two diagnostic strategies have been validated in large scale outcome studies. Both rely on a sequential combination of the above mentioned instruments and have been able to safely manage more than 90% of patients without a pulmonary angiogram. The 3-month venous thromboembolic risk in patients without pulmonary embolism and, hence, not anticoagulated was less than 1% in both studies. In the absence of a formal comparison of their respective cost-effectiveness, choosing between these strategies may rest on local preferences or logistics. Finally, spiral CT scan is highly promising and may considerably modify the diagnostic workup of pulmonary embolism in the near future. However, it is insufficiently validated and its position in a rational algorithm for diagnosing pulmonary embolism is not yet defined.