This paper argues that debates amongst economists triggered by the Stern Review are partly relevant, focusing on key parameters translating real ethical issues, and partly misplaced in that they do not consider enough other determinants of climate change damages: i) the specifications of the utility function used for the assessments (preference for the environment, preference for smooth growth paths), ii) the interplay between uncertainty and the sequentiality of the decision, and iii) whether the growth engines behind the integrated assessment models can account for transient disequilibrium and sub-optimality. We derive some suggestions for any future research agenda in integrated assessment modelling, whatever the position of the analysts about the relevance of the intertemporal optimisation framework and the Bayesian approach to uncertainty in the climate affair.