Background: Approximately 12% of bound blood calcium is linked to various anions including phosphate. In patients with end-stage kidney disease (ESKD), serum phosphate is highly variable. We propose that establishing a formula to calculate albumin- and phosphate-corrected total calcium would be more appropriate to estimate free calcium in ESKD patients.Methods: In 82 haemodialysis patients, serum ionized calcium (Caion) and pH were measured by blood gas analyser with ion-selective electrodes at the point-of-care, while bicarbonate, phosphate, albumin, magnesium and total calcium (Catot) were measured at the central laboratory. Linear regression analysis of measured variables was used to best fit adjusted calcium versus Caion.Results: The most parsimonious multiple linear regression model (r2 = 0.81) of variables associated with Caion included Catot (coeff 0.820, P <0.0001), albumin (coeff −0.016, P <0.0001) and phosphate (coeff −0.063, P <0.002). Modelling of available variables yielded the following equation to adjust calcium for albumin and phosphate: CaalbPh = Catot + (0.015 × (40 − [albumin]) + 0.07 × (1.5 − [phosphate])). At an ambient albumin of 40 g/L, CaalbPh would be 0.07 mmol/L lower than Catot for every mmol/L of phosphate. In vitro data using three different albumin levels and increasing phosphate concentrations demonstrated this relationship, with the slope of the phosphate effect being stronger at lower albumin concentrations.Conclusion: Because guidelines recommendations indicate that corrected serum calcium should be maintained within the normal range in ESKD patients, inclusion of phosphate to correct Catot in these patients may have clinical implications on the choice of phosphate binders and the prescription of vitamin D or calcimimetic agents.