The determination of blood recirculation using blood urea nitrogen (BUN) measurements in hemodialysis patients is a standard technique. The accuracy and reproducibility of these calculations have never been determined. Two pairs of recirculation studies (study A and study B) were performed in 13 patients during a single dialysis treatment. Blood samples were analyzed for BUN and recirculation was calculated. The first recirculation study (study A) was performed within 1 hour of the initiation of dialysis, with a duplicate test of recirculation performed within 15 minutes. In study B, the dialyzer blood lines were reversed in an attempt to enhance blood recirculation. After 15 minutes, duplicate tests of recirculation were again performed. Calculated recirculations before the line reversal (study A) ranged from -3.3% to 11.9% in the first test and -2.9% to 12.2% in the second test. In study A, there was no correlation (P > 0.05, r = 0.09) between the first and second calculated recirculations. In study B, an increase in recirculation was observed. Calculated recirculations ranged from 16.3% to 53.5% for the first test and 5.4% to 58.1% for the second test. A significant relationship was observed in the calculated recirculation in study B (P < 0.05, r = 0.81). The results from the present study show that the use of BUN measurements may not provide a consistent indicator of access recirculation in a patient with a low recirculation. This lack of consistency should be considered when determining further clinical treatment.