Measurements of dialysate sodium are used to estimate peritoneal dialysis sodium losses and sodium sieving, a measure of hydraulic permeability of the peritoneum. Peritoneal dialysates differ from serum samples in terms of pH, osmolality, protein and glucose concentration. We wished to determine whether these factors affected sodium measurement. Dialysate samples were taken from 52 consecutive peritoneal dialysis patients attending for a standard peritoneal dialysis equilibrium test (PET), 20 with standard lactate dialysate and 32 with neutral pH dialysate and sodium was measured by both flame photometry and indirect ion selective electrode (ISE). Sodium measured by ISE consistently overestimated that measured by flame photometer, mean bias 1.5 mmol/L (95% confidence limits 1.2 to 1.8), P < 0.001. Sodium was lower in fresh neutral pH dialysates by both methods - flame 125.3 ± 1.17 vs. 131.6 ± 0.39 mmol/L, than standard lactate dialysates ISE 127.4 ± 1.05 vs 132.7 ± 0.27 mmol/L, P < 0.001. Glucose was higher in fresh neutral pH dialysates 122.7 ± 1.1 vs. standard lactate dialysates 116.7 ± 0.4 mmol/L, P < 0.001. On multiple regression analysis, only glucose was found to be an independent factor for sodium measurement, F = 14.78, β = -0.0851, SEM 0.022, 95% confidence limits -1.28 to -0.042. In this study there was a small but consistent difference between sodium measurements by ISE and flame photometry during the PET. Sodium measurements by either method appeared to be affected by hypertonic dialysates, but there were differences with pH. This may potentially lead to errors in both overestimating peritoneal sodium losses and the proportion of patients with ultrafiltration failure due to loss of sodium sieving.