Impedance-based fault location algorithms require positive- and zero-sequence line impedance to determine the distance to fault. When solving for these line constants in a multi-grounded transmission system, power system analysis programs use Carson's model, modified Carson's model or Deri model to account for earth current return. The study is motivated by the fact that line constants calculated using each of these different approaches are not exactly identical. Since transmission lines traverse long distances, it was suspected that these seemingly minor differences can affect accuracy of fault location estimates. Therefore this paper aims to study the impact of line constants calculated using different earth return models on impedance-based fault location algorithms. Analysis was conducted on four actual fault events. For a fault located 14.37 miles from substation, variation between location estimates using each of the three earth models was observed to be within 1%. The paper therefore recommends using any of the three earth models for fault location without significant variation in accuracy.