Pleuritic chest pain is a frequent complaint in patients coming to the emergency room, but the proportion of such patients with pulmonary embolism is uncertain. In a prospective study, we evaluated the diagnostic outcomes in 173 consecutive patients who came to the emergency room with pleuritic chest pain. Pulmonary embolism, as demonstrated by angiography or autopsy, was present in 36 (21%). The need for objective testing is clearly indicated by our finding that the sensitivity (85%) and specificity (37%) of predetermined clinical variables for pulmonary embolism were insufficient to allow a definitive treatment decision. Optimal sensitivity and specificity are obtained by using pulmonary angiography in combination with lung scanning. The proportion of patients requiring angiography is substantially reduced, from 43% to 26%, without significant loss of accuracy, if ventilation imaging and impedance plethysmography are used together with perfusion scanning.