We studied the diagnostic accuracy of computed tomography coronary angiography (CTCA) for the diagnosis of significant coronary artery disease (CAD) compared to conventional coronary angiography (CCA) in patients with primary cardiac tumors. Thirty-eight consecutive patients with primary cardiac tumors (27 females, 11 males; mean age 56+/-6 years, range 32-86 years) underwent dual-source CTCA and CCA. Significant stenosis was defined as diameter reduction >50%. CCA served as the standard of reference. The prevalence of significant CAD in the study population was 8% (3/38 patients). Five of 544 segments (0.9%) in 1/38 patients (2.6%) was considered of non-diagnostic image quality on CTCA because of motion artifacts. In a segment-based analysis taking not-evaluative segments as false-positive, sensitivity, specificity, positive and negative predictive value of CT was 100%, 99%, 70%, and 100%, respectively. Preoperative CCA could have been avoided in 95% (36/38) of the patients and CCA would have confirmed the CTCA diagnosis in 5% (2/38) of the patients. Our results indicate that CTCA provides a high diagnostic performance for diagnosing significant CAD in patients with primary cardiac tumors. CTCA may thus be used as a filter test prior to surgery.