Abstract. Inferred Effective Climate Sensitivity (ECSinf) is estimated using a method combining radiative forcing (RF) time series and several series of observed ocean heat content (OHC) and near-surface temperature change in a Bayesian-15 framework using a simple energy balance model and a stochastic model. The model is updated compared to our previous analysis by using recent forcing estimates from IPCC, including OHC data for the deep ocean, and extending the time series to 2014. The mean value of the estimated ECSinf is 2.0°C, with a 90% credible interval of 1.2-3.1°C. The mean estimate has recently been shown to be consistent with the higher values for the equilibrium climate sensitivity estimated by climate models. We show a strong sensitivity of the estimated ECSinf to the choice of a priori RF time series, excluding pre-1950 20 data and the treatment of OHC data. Sensitivity analysis performed by merging the upper (0-700m) and the deep ocean OHC or using only one OHC data set (instead of four in the main analysis), both give an enhancement the mean ECSinf by about 50% from our best estimate.