Recent environmental economics literature regarding global problems, like for example those associated with greenhouse gases, focuses on the emergence of international co-operation and underlying incentives. International co-operation among countries is necessary in the case of global pollution, since outcomes related to laissez-faire equilibrium are inefficient, because each country chooses its pollution behaviour ignoring the cost imposed on other countries as a result of its behaviour. Voluntary co-operation among sovereign countries has been analysed in terms of agreements of individual countries or sub-groups of countries to reduce emissions. Various models have been developed, in which a group of countries seeks to expand the agreement by self-financing welfare transfers. This implies side payments to other countries in order to induce them to restrict emissions to some desired level, so that a stable coalition is formed. This paper provides an empirical investigation related to the possibility of international agreements with respect to CO2 emissions. Using data on GDP and CO2 emissions for five groups of countries - Europe, Latin America, the United States and Canada, India, and China - benefit functions relating GDP to CO2 and damage functions from the accumulation of CO2 emissions are estimated. Benefit-cost ratios for changes in CO2 emissions are estimated for the countries in each group; they are used as indicators of whether co-operation is possible.