This article proposes a sequential game methodology to analyse the long-term climate policy prospects. Players can sequentially choose the best policy, among a series of policy options, while reacting to past moves of the other players. In order to illustrate the game methodology, a numerical optimisation exercise is made, based on a simple integrated assessment model. The non-cooperative equilibrium arising from a five-stage sequential game with two large players (Annex B and non-Annex B regions), which tries to replicate the Kyoto and beyond Kyoto scenarios, is studied.