Objective To investigate the efficacy and safety of balloon-assisted catheter directed thrombolysis (CDT) for acute lower extremity deep vein thrombosis (DVT).Methods From September 2008 to February 2011,94 patients with acute lower extremity DVT were admitted.The cases in early stage were treated by CDT (Group A,n =50),and the cases in late stage were treated by balloon-assisted CDT ( Group B,n =44).The clinical data of these patients were retrospectively analyzed.The circumference difference between normal and affected limbs,scores of venous patency,and rates of venous patency were recorded for judging the efficacy.The total dose of urokinase and retention time of infusion catheter was compared between the two groups.The incidence of pulmonary embolism and bleeding were used to judge the safety of treatment.The venous patency was followed up by ultrasound or/and venography.Measurement data with normal distribution was described by mean + standard,and was analyzed using T test.Measurement data with non-normal distribution was described by M ( QL,QU ),QL =P25,QU =P75,and was analyzed using Wilcoxon' s test.Categorical variable data was analyzed using Chi-Square test Results The prior treatment circunfference difference between normal and affectéd limbs were (5.37 ±1.97) cm (thigh) & (4.14 ± 1.57) cm (calf) in Group A and (5.41±2.22) cm (thigh) & (4.05 ±1.61) cm (calf) in Group B ; and the difference between the groups was insignificant ( thigh:t =- 0.113,P=0.910; calf:t =0.288,P =0.774).The post treatment correspondences were:(2.96 ± 1.10) cm (thigh) & ( 1.93 ± 0.84 ) cm (calf) in Group A and ( 1.78 ± 1.40) cm ( thigh ) & ( 1.41± 1.17 ) cm (calf) in Group B; the difference between the groups was significant (thigh:t =4.66,P ＜0.0001; calf:t =2.548,P =0.012 ).The prior treatment score of venous patency was 9 (8,10) in Group A and 8.3(7,10) in Group B without significant difference (Z =- 1.5172,P =0.1292).The post treatment score of venous patency was 3.5 ( 2,5 ) in Group A and 0 ( 0,1) in Group B with significant difference ( Z =-5.7702,P ＜0.01).The rate of venous patency after the treatment was 55.0％ (42.3％,72.4％ ) in Group A and 100％ (88.5％,100％ ) in Group B,with significant difference ( Z =4.9148,P ＜ 0.01).The total dose of urokinase used in the treatment was 5.950 ( 5.525,7.225 ) × 106U in Group A and 4.100 (3.600,5.050) × 106U in Group B with significant difference (Z =-6.0133,P ＜ 0.01).The retention time of perfusion catheter was 10 (9,12) d in Group A and 6 (5,7) d in Group B with significant difference ( Z =- 8.0358,P ＜ 0.01).No symptomatic pulmonary embolism occurred in both groups during the treatment and follow-up period.The rate of bleeding complication was 38.0％ (19/50) in Group A and 22.3％ (10/44) in Group B,without significant difference (x2 =2.5590,P =0.1097 ).The removal rate of optional filter was 88.37％ (38/43) in Group A and 100％ (39/39) in Group B,with significant difference ( x2 =4.829,P =0.028 ).The rate of venous patency at the last follow-up point was 50.0％ (44.4％,59.2％ ) in Group A,and 95.4％ (83.6％,100％ ) in Group B,with significant difference (Z =- 3.2721,P =0.0011).Conclusions Balloon-assisted CDT was a promising treatment for acute lower-extremity DVT.It improved the effect of thrombolysis and reduced the dosage of urokinase,and did not increase the risk of pulmonary embolism.