This study examines how the design of international agreements on greenhouse gas emissions might affect the incentives for countries to develop technology to address emissions. The authors develop a model where:technology spillovers, both within countries and between countries, are occurring causing the incentive for private firms to invest in Research and Development (R&D) to be lowthese technology externalities within each country are corrected through a domestic subsidy of R&D investmentsno instrument to correct for these international externalities is in place via some kind of international agreement or policy.The model is used to test various types of international intervention to determine which would achieve the best results.The authors find that:With an international agreement controlling abatements directly through emission quotas, the equilibrium R&D subsidy is lower that the socially optimal subsidyThe equilibrium subsidy is even lower if the climate agreement does not specify emission levels directly, but instead imposes a common carbon taxSocial costs are higher under a tax agreement than under a quota agreement. Moreover, for a reasonable assumption on the abatement cost function, R&D investments and abatement levels are lower under a tax agreement than under a quota agreementTotal emissions may be higher or lower in a second-best optimal quota agreement than in the first-best optimum.