An increasingly common type of environmental policy instrument limits the carbon intensity of transportation and electricity markets. In order to extend the policy's scope beyond point-of-use emissions, regulators assign each competing fuel an emission intensity rating for use in calculating compliance. I show that welfare-maximizing ratings do not generally coincide with the best estimates of actual emissions. In fact, the regulator can achieve a higher level of welfare by manipulating the emission ratings than by manipulating the level of the standard. Moreover, a fuel's optimal rating can actually decrease when its estimated emission intensity increases. Numerical simulations of the California Low-Carbon Fuel Standard suggest that when recent scientific information suggested greater emissions from conventional ethanol, regulators should have lowered ethanol's rating (making it appear less emission-intensive) so that the fuel market would clear with a lower quantity.