Climate mitigation policy should be imposed over a long period, and spur development of new technologies in order to make stabilization of green house gas concentrations economically feasible. The government may announce current and future policy packages that stimulate current R&D in climate-friendly technologies. However, once climate-friendly technologies have been developed, the government may have no incentive to implement the pre-announced future policies, that is, there may be a time inconsistency problem. We show that if the government can optimally subsidize R&D today, there is no time inconsistency problem. Thus, lack of commitment is not an argument for higher current R&D subsidies. If the offered R&D subsidy is lower than the optimal subsidy, the current (sub-game perfect) climate tax should exceed the first-best climate tax.