Abnormalities in calcium concentration are frequent in patients receiving dialysis therapy. Most cases of both hypo- and hypercalcemia are mild and asymptomatic. There is concern, however, that, on the one hand, hypocalcemia can drive hyperparathyroidism and eventually lead to gland hypertrophy and autonomous function. Hypercalcemia, on the other hand, can be associated with increased extraosseous calcium and phosphate deposition leading to vascular calcification with an attendant mortality and morbidity. Calcium exists in three main forms in the blood: the physiologically active free or ionized fraction (terms often used interchangeably), a protein bound fraction, and a fraction complexed to other anions. Although the ionized calcium can readily be measured using ion-specific electrodes, it is the total calcium that is most commonly measured because of sample handling and cost concerns. As it is the free or ionized form that is biologically active (and therefore of most relevance), a number of adjustment formulae have been derived to "correct" the total calcium for changes in albumin, protein, and complexing ion concentrations. These formulae show good statistical correlation with measured ionized calcium in populations studied as a whole, but are generally poor predictors of true ionized hypo- or hypercalcemia in individual patients. International guideline committees in nephrology recommend frequent assessment of calcium levels in dialysis patients and recommend that these levels be kept within the normal reference range. These guidelines are less clear on which measurement of calcium should be used to guide clinical decision making. This review examines the merits of making any adjustment to the total calcium measurement, and suggests when it is appropriate to measure the ionized or free calcium.