OBJECTIVE Pulse-spray pharmacomechanical thrombolysis is an evolving method for the treatment of vascular occlusions in which a highly concentrated fibrinolytic agent is injected as a high-pressure spray directly into thrombus. The purpose of this retrospective study was to analyze our long-term experience with this technique for the treatment of clotted hemodialysis grafts and to compare the efficacy and safety of the original and current methods. METHODS Over 6 years, 284 cases of dialysis graft thrombosis were considered suitable for treatment with pulse-spray thrombolysis. The original technique involved the injection of highly concentrated urokinase directly into a clot through two crisscross catheters with multiple side holes. The current technique includes early fragmentation of residual clot with a balloon catheter, intrathrombic injection of heparin, mechanical treatment of a lysis-resistant clot at the arterial anastomosis, and routine administration of aspirin. After thrombolysis, underlying obstructions were treated with balloon angioplasty, atherectomy, or stents. The technical success, immediate clinical success, and frequency of complications for the entire population were analyzed. In addition, the results for 36 cases treated with the original technique were compared with the results for 37 recent cases treated with the current technique. RESULTS Of 284 cases considered suitable for treatment, thrombolysis was not done in eight cases because the venous anastomosis could not be crossed. Thrombolysis was discontinued in two cases because of extravasation of contrast material. The technical success for all grafts considered for treatment was 96%; 92% of treated grafts remained patent for at least 24 hr. Major complications occurred in 1% of cases, and minor complications occurred in 9% of cases. The clinical efficacies of the original and current techniques were 86% and 92%, respectively. The mean thrombolytic agent infusion time was reduced from 44 +/- 20 min to 23 +/- 13 min (p < .001). The overall procedure time for the recently treated subgroup was 67 +/- 26 min. There was no significant difference in the frequencies of major and minor complications between the treatment subgroups. CONCLUSIONS Pulse-spray pharmacomechanical thrombolysis is a reliable, rapid, and safe method for recanalization of occluded dialysis grafts. The current technique has been proven as safe and effective as the original technique but offers the advantage of a significant reduction in the time required for the infusion of thrombolytic agent.