The widespread use of ion-selective electrode causes the reference range of the anion gap (AG) to be lowered from 8-16 to 3-11 mmol/l. The use of the outdated reference range (8-16 mmol/l) leads to the misinterpretation of the value of the anion gap. To interpret the anion gap accurately, one must use an analyzer-specific reference range. This study established the reference ranges of the electrolyte and anion gap in four ion-selective electrode analyzers. We collected clotted and lithium-heparinized blood from 124 healthy volunteers. We determined the electrolyte in the Beckman E4A (serum), Beckman Synchron CX5 (serum), and Nova CRT (serum and plasma). The anion gap was calculated from the formula: [Na(+)-(Cl(-)+HCO3(-))]. Blood sodium, potassium and bicarbonate were determined using the Nova Stat Profile Ultra. We used the plasma chloride from the Nova CRT to calculate the value of the anion gap in the Nova Stat Profile Ultra. We established the reference ranges using the non-parametric percentile estimation method. Accuracy and precision of the electrolyte performances obtained from all analyzers were acceptable. Reference values of serum and plasma sodium, potassium, and chloride were similar in all analyzers. The value of blood sodium obtained from the Nova Stat Profile Ultra was slightly higher than the values for the serum and plasma sodium obtained from the other analyzers. The bicarbonate ranges obtained from the Nova analyzers were higher than the values obtained from the Beckman analyzers. For the anion gap, the reference ranges in this study were low but similar to other studies (3-11 mmol/l) using ion-selective electrode. However, our reference ranges were lower than the previous reference ranges obtained from the continuous-flow analyzer (8-16 or 9-18 mmol/l) incorporated with flame photometry and colorimetry techniques.