One of the largest sources of global greenhouse gas emissions can be  addressed through conservation of  tropical forests by channeling funds to developing countries at a cost-savings for developed countries. However, questions remain to be resolved in negotiating a system for including reduced emissions from deforestation and forest degradation (REDD) in a post-Kyoto climate treaty. The approach to determine national baselines, or reference levels, for quantifying REDD has emerged as central to negotiations over a REDD mechanism in a post-Kyoto policy framework. The baseline approach is critical to the success of a REDD mechanism because it affects the quantity, credibility, and equity of credits generated from efforts to reduce forest carbon emissions. We compared outcomes of seven proposed baseline approaches as a function of country circumstances, using a retrospective analysis of  FAO-FRA data on forest carbon emissions from deforestation. Depending upon the baseline approach used, the total credited emissions avoided ranged over two orders of magnitude for  the same quantity of  actual emissions reductions. There was also a wide range in  the relative distribution of  credits generated among the five country types we identified. Outcomes were especially variable for countries with high remaining forest and low rates of deforestation (HFLD). We suggest that the most credible approaches measure emissions avoided with respect to a business-as-usual baseline scenario linked to historic emissions data, and allow limited adjustments based on forest carbon stocks.