Hypoproteinaemia may lead to spuriously high electrolyte values using indirect ion-selective electrodes (ISE) compared to direct ISEs. This study evaluates the impact on electrolyte status assessment of direct compared to indirect ISE sodium and potassium measurements in samples from critically ill patients who have a high prevalence of hypoproteinaemia. Serum sodium and potassium measurements were compared using indirect and direct ISE in 190 samples received from critical care units over a three-week period. Serum sodium and potassium measurements were higher (P < 0.0001) using indirect ISE (140.0 +/- 5.0 and 4.5 +/- 0.6, respectively) compared to direct ISE (136.5 +/- 5.2 and 4.5 +/- 0.6, respectively). The calculated difference between indirect and direct ISE values for sodium increased as total protein concentration decreased (Y = 7.2-0.07X, 95% CI slope -0.1 to -0.05, P<0.0001, r2 = 0.14). Hypoproteinaemia was present in 85% of samples. Indirect ISE, compared to direct ISE, misclassified 28% of samples as pseudonormonatraemia (19%), pseudohypernatraemia (8%), pseudonormokalaemia (0.8%) and pseudohyperkalaemia (0.4%). Hypoproteinaemia is common in critically ill patients and this may lead to spuriously high indirect ISE electrolyte measurements, resulting in significant misclassification of electrolyte (particularly sodium) status. In such patients, direct ISE (as employed in point-of-care testing) offers more accurate and consistent electrolyte results than does indirect ISE (commonly used in major laboratory analysers).