Percutaneous transluminal angioplasty (PTA) is a possible treatment for stenosis. This study aimed to verify the impact of a vascular access (VA) surveillance protocol, based on the detection of functional changes and their correction by a new PTA method for VA performed under color Doppler ultrasonography (CDU) guidance. We divided the patients into two groups: group A, before May 1999 (retrospective study) without the surveillance protocol, and group B, from 1 May 1999 to January 2001 (prospective study) with the surveillance protocol. Access blood flow (Qa) was assessed every 4 weeks by ultrasound velocity dilution. In cases of a reduction of >or=35% from the baseline value, VA was examined using CDU: if a stenosis >50% was detected, angioplasty was performed. In cases of Qa reduction <35% we continued monitoring. By Coxs multivariate analyses, only the use of PTA with or without stenting reduced the relative risk of thrombosis by 64% during the follow-up (p=0.017 confidence intervals 88%-15%) in group B patients. Secondary patency was 80% for VA in which we performed PTA with or without stenting at 18 months, and 58% at 18 months in which we did not perform PTA. Our data show how PTA under CDU is useful to maintain and to improve graft patency. This PTA under CDU guidance allows patients to avoid surgical intervention, hospitalization, and adverse reactions to contrast media and exposure to ionizing radiation, with reduced cost and with better graft survival.