Although monitoring of vascular accesses by physical examination is nearly as sensitive as surveillance measurements by vascular access pressure when performed by examiners, the frequency of examinations is limited by time. We developed intravascular access pressure surveillance as a surrogate to physical examination. Using real-time data from hemodialysis machines, we derived intravascular access pressure ratios for each dialytic procedure. An automated, noninvasive surveillance algorithm that generated a “warning” list of patients at risk for thrombosis was formulated. We hypothesized that this algorithm would reduce access thrombosis frequency. We designed a study comparing thrombosis rates during a baseline 6-month interval to three subsequent 6-month periods of active surveillance. Referrals for interventions during this 18-month period were based on persistently abnormal elevated vascular access pressure ratio tests (VAPRT) >0.55. Thrombosis rates declined progressively for arteriovenous grafts (AVG) during the intervention period compared with the baseline period. Arteriovenous fistula (AVF) thrombosis rates decreased during postintervention months 13–18 during employment of the VAPRT. We conclude that use of VAPRT can reduce thrombosis rates in vascular accesses, and the magnitude of the effect is larger and more consistent in arteriovenous grafts (AVGs) than autologous AVFs.