Abstract Purpose To evaluate inferior vena cava (IVC) venograms (ie, cavograms) before filter retrieval to determine the incidence and volume of filter thrombus relative to filter dwell time and evaluate subsequent changes in thrombus volume with additional anticoagulation. Materials and Methods IVC filter retrieval attempts between December 2002 and June 2010 were retrospectively reviewed to determine the incidence of filter thrombus and estimate thrombus volume on a preretrieval cavogram. Correlation between filter dwell times (assessed at 30-d intervals) and incidence and volume of thrombus was assessed. Follow-up images and management of filters with thrombus that were not initially removed were analyzed. Results A total of 463 retrieval attempts were performed in 440 patients, with a mean filter dwell time of 95 days ± 145 (SD; range, 0–1,762 d). Thirty (6.5%) had filter thrombus on initial cavograms, with a mean thrombus volume of 2.8 cm 3 ± 7.3 (range, 0.04–40.02 cm 3). Incidence rate and estimated thrombus volume were highest in the 0–30-day dwell interval (8.0% and 6.3 cm 3, respectively) and decreased at subsequent time intervals. On linear regression analysis, incidence of filter thrombus was inversely related to dwell time ( P < .05; correlation coefficient, −0.86). Seven patients with thrombus underwent additional anticoagulation for a mean of 48 days ± 25 (range, 14–90 d); thrombus resolved completely in five (71%) and partially in one (14%), and increased in one (14%). Conclusions The incidence of filter thrombus at the time of filter retrieval appears to decrease with dwell time. If thrombus is detected, an additional period of anticoagulation is likely to reduce the thrombus burden and facilitate later retrieval.