This study employs a system GMM dynamic regression model and examines how the rates of deforestation vary by the level of per capita income and other factors for 21 middle/upper income African countries and 9 East Asian countries using panel data from 1990 to 2016. The East Asian sub-sample serves as a benchmark against which to contrast the middle and upper income African sub-sample. Accordingly, the Environmental Kuznets Curve (EKC) hypothesis has been tested and the results demonstrate there is indeed an inverted U-shaped curve between per capita income and deforestation rate for East Asian and middle/upper-income African countries, implying that deforestation initially increases with income level rises, but starts to decline at higher levels of development. Once the EKC confirmed, the turning points (the minimum threshold income per capita) that signify a decrease in the rate of deforestation for East Asian and middle/upper income African countries are estimated to be $2785 and $2307, respectively. In addition to per capita income, other underlying factors of deforestation have been identified. In addition to the GMM estimation method, FE and RE estimation methods are also used to verify the consistency and robustness of the empirical results.