The purpose of the study was to assess the clinical safety and efficacy of the "Recovery(TM)" (Bard) inferior vena cava (IVC) filter. We retrospectively evaluated the clinical and imaging data of patients who had a "Recovery(TM)" IVC filter placed between January 2003 and December 2004 in our institution. The clinical presentation, indications, and procedure-related complications during placement and retrieval were evaluated. Follow-up computed tomography (CT) examinations of the abdomen and chest were evaluated for filter-related complications and pulmonary embolism (PE), respectively. "Recovery" filters were placed in 96 patients (72 males and 24 females; age range: 16-87 years; mean: 46 years). Twenty-four patients presented with PE, 13 with deep vein thrombosis (DVT) and 2 with both PE and DVT. The remaining 57 patients had no symptoms of thromboembolism. Indications for filter placement included contraindication to anticoagulation (n = 27), complication of anticoagulation (n = 3), failure of anticoagulation (n = 5), and prophylaxis (n = 61). The device was successfully deployed in the infrarenal (n = 95) or suprarenal (n = 1) IVC through a femoral vein approach. Retrieval was attempted in 11 patients after a mean period of 117 days (range: 24-426). The filter was successfully removed in nine patients (82%). Failure of retrieval was due to technical difficulty (n = 1) and the presence of thrombus in the filter (n = 1). One of the nine patients who had the filter removed developed IVC thrombus after retrieval and another had an intimal tear of the IVC. Follow-up abdominal CT (n = 40) at a mean of 80 days (range: 1-513) showed penetration of the IVC by the filter arms in 11, of which 3 had fracture of filter components. In one patient, a broken arm migrated into the pancreas. Asymmetric deployment of the filter legs was seen in 12 patients and thrombus within the filter in 2 patients. No filter migration or caval occlusion was encountered. Follow-up chest CT (n = 27) at a mean of 63 days (range: 1-386) showed PE in one patient (3%). During clinical follow-up, 12 of 96 patients developed symptoms of PE and only 1 of the 12 had PE on CT. There was no fatal pulmonary embolism in our group of patients following "Recovery" filter placement. However, the current version of the filter is associated with structure weakness, a high incidence of IVC wall penetration, and asymmetric deployment of the filter legs.