Blood flow in peripheral arteriovenous fistulae and grafts as used for hemodialysis access can be derived from measurements of the amount of access recirculation induced by reversing the dialysis blood lines and a knowledge of the dialyzer blood flow rate. The Hemodynamic Monitor (HDM; GAMBRO Healthcare, Lakewood, CO) is a device that uses magnetic principles to accurately and precisely measure access recirculation during hemodialysis. The measurement is based on differential conductivity between arterial and venous blood flows in the dialysis blood tubing sets following the injection of hypertonic saline into the venous line as a conductivity tracer. Clinical studies were performed on 41 patients from two centers who had arteriovenous fistulae (25 patients) or Goretex grafts (16 patients; W.L. Gore & Associates, Flagstaff, AZ); each patient was studied on two successive dialysis days under variable conditions of dialyzer blood flow, and multiple measurements were made according to a standard protocol. The protocol involved temporarily reversing the arterial and venous lines, then performing an HDM recirculation test and recording the result along with the dialyzer blood flow rate as per the machine blood pump setting. The access blood flow rates measured 1,125+/-581 mL/min (mean+/-SD) on day 1 and 1,140+/-680 mL/ min on day 2 (P > 0.05 [NS]), with an absolute range of 221 to 3,118 mL/min. These flow rates are similar to those measured by other techniques. There was an excellent correlation between access blood flow rates measured in individual patients on days 1 and 2, even in a subset of 13 patients who had the dialyzer blood flow rates altered by > or =100 mL/min, suggesting the independence of access from dialyzer blood flow rates. Analysis of repeated measurements of access blood flow under identical conditions showed a characteristic standard deviation from the mean across the patient population of 7.89%, indicating that the HDM results are repeatable in clinical application. The influence of the measured access blood flow on the outcome of that access was determined after an 8-month follow-up period. Of the 41 accesses, nine were lost to clotting; seven of 14 that had initial blood flow rates less than 750 mL/min clotted, while only two of 27 with flow rates greater than 750 mL/min subsequently clotted (P = 0.005). The data show that the HDM can provide clinically important information on access blood flow.