Computation of distance to fault on a transmission line is affected by: assumptions in modelling, factors reflecting system operating conditions, and measurement system characteristics. All these factors are subject to many sources of uncertainty including parameter setting errors, measurement errors, as well as absence of information and incomplete modelling of a system under fault conditions. These impose a limit on the confidence of fault location. In this paper we propose variance-based sensitivity measures for evaluation of fault location techniques. The main goal of the evaluation is to determine factors and their interactions that contribute to the fault locator output variability. In practice, such analysis can help in: selection of the optimal fault locator for a specific application, calibration process and building confidence in a fault location result. The paper concludes with an application example which demonstrates use of the proposed methodology in testing and comparing two commonly used fault location algorithms.