The objective of this paper is to propose a method for evaluating fault identifiability in linear discrete time-varying systems. The basic idea of this method is to quantify the difficulty in estimating an unknown value from an observation for the constructed moving window time-varying model. The analysis of the connection among fault detectability, fault isolability, and fault identifiability is presented in terms of the corresponding measures, distinguishability, and estimability, respectively. Furthermore, factors that influence the performances are analyzed, and an approach for selecting a better system configuration is proposed based on the Fisher information matrix. Finally, the effectiveness of the proposed method in terms of qualitative and quantitative analyses is illustrated through two examples.