Surgery is the traditional treatment for juxta-anastomotic stenoses in forearm arteriovenous fistulas (AVF), but percutaneous transluminal angioplasty (PTA) is a suitable alternative. No prospective comparative trials between the two have been reported to date, however. A retrospective analysis of prospectively, concurrently collected data was performed to compare the outcome and cost of surgery and PTA in the preemptive repair of juxta-anastomotic stenosis in lower forearm AVF. Sixty-four AVF with >50% venous juxta-anastomotic stenosis were considered: 21 were treated surgically (11 proximal neo-anastomosis and 10 polytetrafluoroethylene interposition graft) and 43 by PTA. After treatment, AVF were monitored by quarterly ultrasound dilution access blood flow measurement. End points were restenosis and procedure failure rate (re-intervention by another technique or access loss), and determinants were analyzed using Cox hazard model. Initial procedural success was 100% for surgery and 95% for PTA (P = 0.539). Restenosis rate was 0.168 and 0.519 events/AVF-year for surgery and PTA, respectively (P = 0.009). The type of procedure was the only variable that was significantly associated with restenosis, the adjusted relative risk being 2.77-fold higher (95% confidence interval 1.07 to 7.17; P = 0.036) after PTA than surgery. The procedure failure rate was 0.110 and 0.097 events/AVF-year for surgery and PTA, respectively (P = 0.736). The cost profile also was similar for the two procedures. This prospective comparative study confirms a higher restenosis rate after PTA than surgery, but with strict surveillance for restenosis, the two procedures show similar assisted primary patency and cost, suggesting that they should be considered equally valid, complementary alternatives in the preemptive treatment of juxta-anastomotic stenosis in forearm AVF.