We report the efficacy of catheter-directed thrombolysis (CDT) for acute deep vein thrombosis. Between January 2003 and August 2004, 20 patients were treated with CDT for occlusive femoral, ilio-femoral and vena caval thrombosis, for less than 2 weeks from onset. Average age was 56.4 years (range 30-78 years), 11 patients were male, and the duration of leg symptoms was 4.4 days (range 1-12 days). Routine temporary inferior vena caval filters were used, and a multi-lumen catheter was inserted from the popliteal vein. Urokinase was used via the catheter by the combination drip infusion method and pulse-spray method. All patients received heparin and stasis of venous flow was prevented with intermittent pneumatic compression. If thrombus remained, mechanical thrombolysis was necessary. Metallic stents were implanted for iliac vein compression syndrome and organized thrombus. Venographic severity score (VS score) and extremity circumference were used to evaluate the effects of treatment. The duration of the treatment was 5.0±0.28 days (range 2-9 days) and the total dosage of urokinase was 1, 025, 000±57, 000 units (range 360, 000-1, 680, 000 unit). One (5%) iliac vein compression syndrome and two (10%) organized thrombi were treated by implanted metallic stents. Giant thrombi was captured by temporary inferior vena caval filters in two patients, but there was no pulmonary embolism. Two patients had thrombophilia, one was antiphospholipid syndrome and one was protein S deficiency. There was an early recurrence in one patient and re-CDT was needed. The VS score deteriorated to 6.2±2.5 (post CDT) significantly (p<0.0001) from 26.2±6.3 (pre CDT). CDT for acute deep vein thrombosis was effective and its early outcome was acceptable.