Because pulmonary embolism is often silent, simple clinical procedures are desirable to identify patients with a low to intermediate probability of pulmonary embolism. Among 19,467 patients managed under general anesthesia, we had one bile tract surgery case and three neurosurgical cases whose silent pulmonary embolism was initially suggested by an increase in the arterial to end-tidal CO2 gradient (from 17 to 27 mmHg) after general anesthesia was induced or their trachea was intubated. During the preoperative assessment, the patients presented no clinical manifestations suggestive of pulmonary embolism. Our initial diagnosis was confirmed by scintigraphy and/or angiography done immediately after the operations. Because capnometry has been shown to be applicable to non-intubated, spontaneously breathing patients, we suggest that measuring the gradient may serve as an additional method for unmasking silent pulmonary embolism in patients at risk or with disturbed consciousness, whether they are scheduled for operations or not.