This paper intends to draw overall lessons from a long-term study on CO2 emission in France, a country with a rather low energy/GDP ration and in which transition to non-fossil based production of electricity has been achieved. More generally, it points out the importance of possible bifurcation effects, and draws methodological and policy implications from these statements. It discusses modelling approaches to the linkages between energy and the rest of economy, and addresses the issue of difficulties in cost assessment analysis in the presence of several baseline scenarios. Finally, it proposes a more encompassing definition of no regrets policies.