This paper provides a comprehensive review of model-based fault diagnosis methods for lithium-ion batteries (LIBs). It begins by classifying battery models into physics-based electrochemical models (EMs) and electrical equivalent circuit models (ECMs). A general state-space representation for a faulty battery model is presented, including the formulation of state vectors and parameter matrices. The paper then discusses various fault mechanisms, such as overcharge/overdischarge faults, connection faults, short circuit faults, and sensor faults. Modeling uncertainties, including modeling errors, measurement noises, aging effects, and measurement outliers, are also addressed. The paper emphasizes the design of observer algorithms, both online and offline, for state estimation and prediction. Finally, it reviews the comprehensive fault diagnosis procedure, which includes fault detection, identification, and estimation, and provides future research directions.This paper provides a comprehensive review of model-based fault diagnosis methods for lithium-ion batteries (LIBs). It begins by classifying battery models into physics-based electrochemical models (EMs) and electrical equivalent circuit models (ECMs). A general state-space representation for a faulty battery model is presented, including the formulation of state vectors and parameter matrices. The paper then discusses various fault mechanisms, such as overcharge/overdischarge faults, connection faults, short circuit faults, and sensor faults. Modeling uncertainties, including modeling errors, measurement noises, aging effects, and measurement outliers, are also addressed. The paper emphasizes the design of observer algorithms, both online and offline, for state estimation and prediction. Finally, it reviews the comprehensive fault diagnosis procedure, which includes fault detection, identification, and estimation, and provides future research directions.