The indication of vena cava filter implantation is controversially discussed. A pure prophylactic indication is increasingly favoured, especially for temporary filter systems without any anamnestic pulmonary embolisms. On the basis of the available literature and our own results a critical analysis of this issue is given. Between 1994 and 1997, we inserted a total of 24 vena cava filters; 21 temporary and 3 permanent filters. In ten patients, the placement of the filter was indicated due to pulmonary embolism and a contra-indication to dose adjusted heparin therapy. Seven additional patients experienced a recurrent pulmonary embolism despite adequate heparin therapy. A prophylactic filter insertion was carried out in seven patients. The temporary vena cava filters were left in place between 7 to 38 days with an average of 19 days. Total implantation time of temporary filters was scheduled until complete mobilisation of the patients, generally in conjunction with an effective dosage of oral anticoagulants. No patient died in connection with the insertion of the filter and no further pulmonary embolisms occurred. One case of inferior vena cava thrombosis occurred in each group of temporary and permanent filters. In one third of the removed filter systems thrombi in the filter were found. Local infections of the catheter and introducer sets were observed in two patients. Moreover, in one case the strut of a temporary filter broke and subsequently dislocated 17 days after insertion. We conclude on the basis of these complication rates that until the results of randomised studies are available the usage of all filter systems should be limited to highly selected cases.