An open circuit voltage-based model for state-of-health estimation of lithium-ion batteries is proposed and validated in this work. It describes the open circuit voltage as a function of the state-of-charge by a polynomial of high degree, with a lumped thermal model to account for the effect of temperature. When applied for practical use, the model requires a prior learning from the initial charging or discharging data for the sake of parameter identification, using e.g. a nonlinear least squares method, but it is undemanding to implement. The study shows that the model is able to estimate the state-of-health of a LiFePO4 cell cycled under conditions where the temperature has fluctuated significantly with a relative error less than 0.45% at most. A short part of a constant current profile is enough for state-of-health estimation, and the effect of size and location of voltage window on the model's accuracy is also studied. In particular, the reason of accuracy change with different voltage windows is explained by incremental capacity analysis. Additionally, the versatility and flexibility of the model to different chemistries and cell designs are demonstrated. •An open circuit voltage-based model is proposed to estimate the SOH of LIBs.•The accuracy of the model is validated by experimental results of an LFP cell.•The effect of size and location of voltage window on the model accuracy is studied.•ICA is used to explain the model accuracy change with different voltage window.•The versatility of the model to different chemistries is demonstrated.