Lithium-ion batteries (LIBs) have become the mainstream power source for battery electric vehicles (BEVs) with relatively superior performance. However, LIBs experience battery aging and performance degradation due to the external environment and internal factors, which should be reflected in the evaluation of the state of health (SOH). Accurately predicting SOH can improve the overall life of the battery and support safe driving in BEVs. At present, while there are many prediction methods for SOH, most are implemented in simulated environments but are challenging to execute in actual industrial production. This review provides a discussion on the aging reasons for LIBs, introduces the SOH prediction method based on the classification framework, and analyzes the key benefits and drawbacks of each method. Finally, the corresponding suggestions and solutions are given in combination with the actual industrial production. •Deficiencies of the existing SOH prediction review are highlighted.•The cause of battery aging is discussed.•The differences and links between SOH and RUL are explained.•The current SOH prediction method based on the classification framework is discussed in detail.•The systematic evaluations and recommendations for the current SOH forecasting method, in conjunction with actual commercial and industrial conditions, are provided.