Background: To investigate the clinical outcomes of indwelling needle-delivered local urokinase thrombolysis, a cost-saving thrombolytic approach, in the treatment of thrombosed arteriovenous grafts (AVGs). Methods: The clinical data of 71 patients with a first episode of thrombosed AVG were analyzed. Among them, 49 patients underwent urokinase thrombolysis and percutaneous transluminal angioplasty (PTA), and 22 patients underwent thrombectomy and PTA. The operation success rate, operation time, perioperative and postoperative complications, primary patency, and secondary patency were compared. Results: Clinical success was achieved in all 71 PTAs. After thrombolysis, 29 patients achieved complete lysis, and 20 patients achieved partial lysis. All patients achieved thrombus clearance after PTA. The operating room usage time was 37.48 ± 5.33 min for the complete thrombolysis group and 41.70 ± 6.16 min for the incomplete thrombolysis group, and these values were not significantly different (p = 0.853). The operating room usage time of the thrombectomy group was 124.73 ± 5.08 min, which was significantly longer than that of the other groups (p < 0.001). The log-rank test indicated no significant difference in the 90-, 180-, and 360-day primary patency and corresponding secondary patency (all p > 0.05). The Cox regression showed that no significant difference was noted in the changing rate of primary or secondary patency among the groups (all p > 0.05). No major complications occurred in any of the groups, and the difference in the complication incidence was not significant (p = 0.336). Conclusions: Indwelling needle-delivered urokinase thrombolysis is an acceptable method for the treatment of thrombosed AVGs. Compared with thrombectomy and PTA, thrombolysis and PTA can significantly shorten the time to use the operating room without reducing primary and secondary patency or increasing complications.