The 'Bali Road Map' of UNFCCC COP-13 calls for sharing lessons learned from demonstration activities that aim to reduce emissions from deforestation and degradation and enhance forest carbon stocks (now known as ‘REDD+'). To develop a feasible yet rigorous strategy for learning from these REDD+ pilots, it is critical to assess previous efforts to evaluate the impacts of 'pre-REDD+' avoided deforestation projects. Further, because REDD+ remains a politically volatile issue, with both critics and supporters pointing to the impacts (or lack thereof) of these pre-REDD+ projects, it is important to critically examine the methods employed to assess those impacts. We review the body of literature that makes claims about the socioeconomic and biophysical impacts of pre-REDD+ projects. We find assessments of outcomes or impacts for only five pre-REDD projects. The design, data collection, and analysis methods for understanding the impacts of pre-REDD+ projects frequently lack rigor. In particular, the counterfactual scenarios for establishing socioeconomic impacts are vague, unscientific, or omitted completely. We conclude that drawing specific lessons from pre-REDD+ projects for the design or evaluation of current REDD+ projects is tenuous. Rigorous project evaluations are challenging, expensive, and time-consuming, but because they are so critical for learning about what works for people and forests, evaluations of current REDD+ projects must use improved methods. In particular, much better care should be taken to construct credible - and where possible, consistent - counterfactuals for both biophysical and socioeconomic outcomes.