The purpose of this study was to compare angiography and angioscopy for the detection of thrombus during coronary interventional procedures. The diagnosis of coronary thrombus has important clinical implications. Angioscopy can directly visualize the coronary luminal surface and may be more accurate than angiography in the diagnosis of thrombus. Angiography and angioscopy were sequentially performed in 75 patients undergoing a variety of interventional cardiology procedures during 117 distinct procedural time points. The angiographic presence of thrombus was defined as a noncalcified filling defect outlined on at least 3 sides by contrast media. The angioscopic presence of thrombus was defined as red material protruding into the lumen (intraluminal thrombus) or adherent to the luminal wall (mural thrombus) that persisted despite flushing. Thrombus was detected on 14 occasions (12.0%) by angiography compared with 48 (41.0%) by angioscopy (p < 0.05). In 4 of the 14 episodes (28.6%) of angiographic filling defects, angioscopy found no evidence of thrombus and provided an alternative explanation for the angiographic filling defect. When angioscopy was used as a reference standard, the sensitivity of thrombus detection by angiography was 20.8%, with a specificity of 94.2% and a predictive value of 71.4%. The sensitivity of angiography for the detection of intraluminal (protruding into the lumen) thrombus was 100% compared with only 10% for mural (adherent to the luminal wall) thrombus (p < 0.05). Angioscopy was significantly more accurate than angiography for detecting coronary thrombus and may be considered an improved reference standard for this diagnosis.