The clinical utility of indirect computed tomography venography (CTV) of the legs, performed after computed tomography pulmonary angiography (CTPA), to evaluate for deep venous thrombosis (DVT) remains a subject of ongoing debate. Visualization of DVT on CTV requires adequate opacification of venous blood. The objective of this study is to measure the Hounsfield unit (HU) clot-to-blood gradient on CTV and CTPA. Secondary analysis of computed tomography images from a prospective clinical trial was performed. All images were interpreted independently by a radiologist. Using the "region of interest" tool feature of the picture archiving and communication system, observers measured the HU of blood and thrombus on CTPA and CTV, using a structured anatomic approach. Interobserver coefficient of variability (CV%) was examined in all disease-positive cases and in a random sample of 25 disease-negative cases. One hundred seventy-nine patients were enrolled; and 23 had pulmonary embolus (PE) without DVT (13%), one had DVT without PE (0.6%), and 18 had both (10%), leaving 137 (77%) who had no venous thromboembolism (VTE). For the 137 patients with no VTE, the mean (+/-standard deviation) HU values measured for contrast-containing blood with no thrombus were as follows: left and right main pulmonary arteries (MPA), 291+/-73 and 291+/-76; left and right common femoral vein (CFV), 92+/-17 and 88+/-19; and left and right popliteal vein (PV), 87+/-20 and 84+/-20. On CTPA, the HU measured from the filling defects interpreted as acute PE was 5+/-20, suggesting a between-group HU gradient of 250 or 471% for MPA (PE-to-blood). On CTV, the HU for DVT was 63, suggesting a gradient of 27 HU or 43% for CFV (DVT-to-blood) and 23 HU or 37% for PV (DVT-to-blood). Interobserver CV% were: left and right MPA, 5+/-4.0% and 7+/-5.4%; left and right CFV, 7+/-6.8% and 7+/-7.3%; left and right PV, 7+/-8.4%, 6+/-8.2%, and 32+/-27.1% for PE thrombus, and 20+/-30.5% for DVT thrombus. CTV imaging produces a very small HU gradient between clot and blood in leg veins. These findings highlight the importance of ancillary radiological findings in assessing presence of DVT on CTV.