The major clinical use of ventilation-perfusion (V/Q) scintigraphy is for the diagnosis of pulmonary embolism (PE). Accurate diagnosis of PE is essential since effective treatment is available but involves some risk to the patient. The scintigraphic characteristics of PE are segmental perfusion defects in lung that is normally ventilated and normal on the radiograph. The inherent shortcoming of perfusion scintigraphy is its lack of specificity. Combining a ventilation study with perfusion imaging improves the diagnostic specificity of lung scintigraphy. Xenon-133 is currently the most commonly used radionuclide for routine ventilation studies; a long washout technique is more sensitive than single-breath imaging when this radionuclide is used. We obtain preperfusion xenon-133 ventilation studies with a 4-min rebreathing equilibrium phase and a long 5-min washout phase to obtain maximum information. It is imperative that V/Q studies be interpreted with a current high quality chest radiograph. Interpretation of V/Q studies for PE is perhaps best done by assigning a probability diagnosis, since rarely is absolute specificity possible. This article details the criteria we use for these probability determinations.