Hemodialysis accesses must supply adequate blood flow to perform hemodialysis and maintain access patency. Access flow (QA) is not measured routinely during hemodialysis. The purpose of this study was to evaluate whether access flow changes during hemodialysis and to determine which factors correlate with QA. The authors measured hemodialysis access flow by ultrasound dilution (QA-T) (Transonic HD01 hemodialysis monitor; Transonic Systems, Inc., Ithaca, NY) and duplex ultrasound, with time-domain correlation (QA-S) (Philips CVI Phillips Medical Systems, Santa Ana, CA) hourly, in 19 patients during hemodialysis. Mean arterial pressure (MAP) (Fresenius automated blood pressure cuff) and cardiac output (CO) (Transonic hemodialysis monitor; Transonic Systems, Inc.) also were measured sequentially. Using duplex ultrasound, access flow was unchanged. However, it fell 132 +/- 137 ml/m (p < 0.05) by ultrasound dilution in hr 4. Cardiac output fell 586 +/- 840 ml/ m (p < 0.05), and MAP fell 11.9 +/- 13.0 mmHg (p < 0.01). There were small positive correlations between CO and QA (correlation coefficient (r) = 0.32, QA-T; r = 0.27, QA-S; p < 0.05), and between CO and MAP (r = 0.35; p < 0.01). In conclusion, access flow, CO, and MAP decreased modestly during hemodialysis. Further studies are necessary to see if access flow is similar off dialysis, and whether in-line access flow measurements can decrease access thrombosis.