Limited data are available about the diagnostic value of D-dimer testing in cancer patients with clinically suspected pulmonary embolism (PE). Therefore, we evaluated i) the safety and clinical usefulness of an ELISA D-dimer test to rule out PE in cancer patients compared with non-cancer patients and ii) whether adopting a higher D-dimer cut-off value might increase the usefulness of D-dimer in cancer patients. We analysed data from two outcome studies which enrolled 1,721 consecutive patients presenting in the emergency department with clinically suspected PE. Presence of an active malignancy was abstracted from the database. All patients underwent a sequential diagnostic work-up including an ELISA D-dimer test and a 3-month followup. Sensitivity and predictive value (NPV) were 100% in both cancer and non-cancer patients. PE was ruled out by a negative D-dimer test in 494/1,554 (32%) patients without cancer, and in 18/164 (11%) patients with a malignancy. At cut-off values varying from 500 to 900 microg/l, the sensitivity was unchanged (100%, 95% CI: 93% to 100%) and the specificity increased from 16% (95% CI: 11% to 24%) to 30% (95% CI: 22% to 39%). The 3-month thromboembolic risk was 0% (95% CI: 0% to 18%) in cancer patients with a negative D-dimer test. ELISA D-dimer appears safe to rule out pulmonary embolism in cancer patients but it is negative in only one of ten patients at the usual cut-off value. Increasing the cut-off value of D-dimer in cancer patients might increase the test's clinical usefulness.