Using a simple model designed for transparency but nonetheless calibrated to support the much-quoted damage estimates of the Stem Review of the Economics of Climate Change, we demonstrate significant sensitivity of those results to assumptions about the pure rate of time preference, the time horizon, and the rates of risk and equity aversion used to compute certainty- and equity-equivalent annuities. Most importantly, we demonstrate enormous sensitivity to presumed constant regional vulnerability and underlying assumptions about adaptive capacity. Manipulation of any of these parameters one at a time across reasonable ranges can diminish damage estimates by as much as 84% or, in the case of extending the time horizon with the Review's low discount rate, increase damage estimates by 900%. We also confirm the usual result that limiting atmospheric concentrations to specific benchmarks above 400 ppm cannot eliminate all damages. Nonetheless, we applaud the Stem Review author team for reconfirming that the climate problem can be approached productively as an economic problem whose solutions can be explored with the tools of decision analysis. (C) 2008 Elsevier Ltd. All rights reserved.