Should "avoided tropical deforestation" be used to alleviate climate change? It is hotly debated today whether the reduction of tropical deforestation should be encouraged by means of a mechanism within the Convention on Climate Change. This mechanism ("avoided deforestation") would benefit developing countries which voluntarily reduce their deforestation rates, thereby generating at least two positive impacts: (i) an increase in the financial resources available to curb tropical deforestation, with expected positive side-effects on biodiversity conservation, the environmental services provided by these forests, and sustainable development as a whole, (ii) improved effectiveness of the global fight against climate change since tropical deforestation contributes extensively to world carbon emissions. Several proposals were designed for such a mechanism, yet their implementation poses significant methodological problems: (i) baselines calculated ex ante lack accuracy due to insufficient knowledge concerning the direct and underlying causes of deforestation, (ii) baselines calculated ex post lack legitimacy as they only refer to past trends, and (iii) uncertainty as to the actual possibility of relating a reduction in the deforestation rate to public policy options in the host country. Drawing lessons from our analysis, we recommend not promoting mechanisms based on financial rewards for an assumed voluntary reduction of national tropical deforestation rates. Two reasons justify our standpoint: the mechanism would probably generate fake reductions ("hot air") and undesirable side-effects would appear that are detailed in this paper. Rather, we encourage industrialized countries to better use already existing multi- and bilateral instruments, which focus on bettering governance deficiencies in countries hosting tropical forests. Besides, "perverse incentives" should be suppressed from public policies in tropical countries. [ABSTRACT FROM AUTHOR]