We elucidate the differences between absolute and intensity-based limits of CO2 emission when there is uncertainty about the future. We demonstrate that the two limits are identical under certainty, and rigorously establish their relative attractiveness under two criteria: preservation of expectations—the minimization of the difference between the actual level and the initial expectation of abatement associated with a one-shot emission target, and temporal stability—the minimization of the variance of abatement due to fluctuations in emissions and GDP over time. Empirical tests of these theoretical propositions indicate that intensity caps are preferable for a broad range of emission reduction commitments. This finding is robust for developing countries, but is more equivocal for developed economies. / Abstract in HTML and technical report in PDF available on the Massachusetts Institute of Technology Joint Program on the Science and Policy of Global Change website (http://mit.edu/globalchange/www/). / ISW was supported by U.S. Department of Energy Office of Science (BER) Grant No. DE-FG02-02ER63484. ADE and JMS were supported by the MIT Joint Program on the Science and Policy of Global Change, funded through a government-industry partnership including U.S. Department of Energy Office of Science (BER) Grant No. DE-FG02-94ER61937, U.S. Environmental Protection Agency Cooperative Agreement No. XA-83042801-0, and a group of corporate sponsors from the U.S. and other countries.