Abstract: The online estimation of battery state-of-health (SOH) is an ever significant issue for the battery management system. Recently, due to its advantages such as model-free and flexibility, data-driven based methods are promising for online SOH estimation. Aiming at the problems of heavy computing burden and difficulty in implementing for microcontroller of the existing battery SOH estimating methods, a novel estimation approach based on the partial charging voltage segment and kernel ridge regression (KRR) for the SOH of lithium-ion batteries was proposed. KRR combines ridge regression with the kernel trick, which thus learns a non-linear function between the partial charging voltage segment and lithium-ion batteries SOH by the respective kernel and the data. The experimental results on two lithium-ion battery degradation datasets show that the proposed method can achieve fast and accurate SOH estimation, which can be further applied to existing BMS only by adopting partial charging voltage curve segments that could be easily obtained in actual working condition.