Purpose To determine if valuable information could be obtained from abdominal computed tomography (CT) performed before insertion of an inferior vena cava (IVC) filter. Materials and Methods A retrospective review was performed on IVC filter insertions with a CT performed before the procedure. Cavagram and CT were compared for renal vein and IVC anatomy, the diameter of the IVC, and the prevalence of iliocaval thrombus. Correlations were assessed among 3 reference standards for measuring the IVC at cavography. Results The mean IVC diameter was 23.0 mm on CT. On cavagram the mean IVC diameter was assessed by using 3 reference standards: 20.7 mm, with the catheter tip as a reference; 26.9 mm, with a radiopaque ruler; and 23.4 mm, by using a lumbar vertebral body. There was good correlation among the 3 measures of IVC diameter (Pearson's r = 0.75, P < .0001) but moderate correlation with CT (r = 0.36–0.56, P < .001). The sensitivity of cavagram for detecting retroaortic and circumaortic renal veins was 40% and 0%, respectively. Nineteen accessory renal veins (12.8%) were not seen by cavagram. Thirteen patients (8.8%) had iliocaval thrombus on cavagram, of which 12 (92.3%) were not previously detected by CT. Conclusions CT is more sensitive than cavagram for detection of renal vein variants and the level of the lowest renal vein. Therefore, if available, the CT should be reviewed before placement of an IVC filter to optimize positioning. Cavagram remains the criterion standard for detection of iliocaval thrombosis and is necessary before IVC filter insertion.