This article investigates efficient policies against global warming in the case of multiple greenhouse gases. In a dynamic optimization model conditions for an efficient combination of greenhouse gases are derived. The model is empirically specified and adapted to a simulation approach. By various simulation runs, the economics of greenhouse gas accumulation are illuminated; and in particular, it is shown that a CO2-policy alone would most likely lead to an allocation far from efficiency. These results indicate, that policy measures against global warming should allow for substituting between different greenhouse gases. Such a policy would mainly affect the agricultural sector because livestock and intensive farming techniques contribute significantly to the emission of greenhouse gases.