The purpose of this study was to review the use, safety, and efficacy of retrievable inferior vena cava (IVC) filters in their first 5 years of availability at our institution. Comparison was made with permanent filters placed in the same period. A retrospective review of IVC filter implantations was performed from September, 1999, to September, 2004, in our department. These included both retrievable and permanent filters. The Recovery nitinol and Günther tulip filters were used as retrievable filters. The frequency of retrievable filter used was calculated. Clinical data and technical data related to filter placement were reviewed. Outcomes, including pulmonary embolism, complications associated with placement, retrieval, or indwelling, were calculated. During the study period, 604 IVC filters were placed. Of these, 97 retrievable filters (16%) were placed in 96 patients. There were 53 Recovery filter and 44 Tulip filter insertions. Subjects were 59 women and 37 men; the mean age was 52 years, with a range of from 18 to 97 years. The placement of retrievable filters increased from 2% in year 1 to 32% in year 5 of the study period. The total implantation time for the permanent group was 145,450 days, with an average of 288 days (range, 33-1811 days). For the retrievable group, the total implantation time was 21,671 days, with an average of 226 days (range, 2-1217 days). Of 29 patients who returned for filter retrieval, the filter was successfully removed in 28. There were 14 of 14 successful Tulip filter retrievals and 14 of 15 successful Recovery filter retrievals. In one patient, after an indwelling period of 39 days, a Recovery nitinol filter could not be removed secondary to a large clot burden within the filter. For the filters that were removed, the mean dwell time was 50 days for the Tulip type and 20 days for the Recovery type. Over the follow-up period there was an overall PE incidence of 1.4% for the permanent group and 1% for the retrieval group. In conclusion, there was an increase in the use of retrievable filters over the study period and an overall increase in the total number of filters implanted. The increased use of these filters appeared to be due to expanded indications predicated by their retrievability. Placement and retrieval of these filters have a low risk of complications, and retrievable filters appeared effective, as there was low rate of clinically significant pulmonary embolism associated with these filters during their indwelling time.